Overview The whole argument, one page
Part one · Competitor analysis Part one, from the top Mandate Market map Players Matrix Retention The gap
Part two · Feature set Part two, from the top Prior art Front door Self-directed Community Clinical Crisis Back office Rollout
Part three · Business model Part three, from the top Revenue models Price anchors Our pricing Build & run cost ROI case The ask
Part four · Regulatory & data compliance Part four, from the top The landscape The minors problem What Sachetana would generate What the sixteen collect When it goes wrong Design implications

MAHE mental health platform  ·  Discovery dossier

Sachetana v2: the case, on one page.

What sixteen competitors ship, the position none of them occupy, what MAHE would need to build to occupy it, and what that would cost and return. This page is the whole argument; the four parts below are where it's sourced.

Prepared forMAHE, discovery discussion
Date13 September 2026
Scope4 parts · 16 competitors · 15 proposed features

Why this matters

MAHE is not short on psychiatric care. Kasturba Hospital, a Department of Psychiatry, and salaried counsellors already exist, already paid for. What's missing is everything between a student noticing something is wrong and that student sitting down in front of one of them. That missing piece is the whole proposal.

13,892
student suicides in India in 2023, the highest count in a decade
NCRB · 8.1% of all suicides
13%
of students use on-campus counselling, though over half report fair or poor mental health
College Pulse, 2022
2–8 wks
typical wait for a campus counselling appointment once a student does ask
US campus benchmark
3.3%
median 30-day retention across 93 mental health apps, the category's central failure
Cross-app analysis

A court order now turns this from goodwill into a compliance obligation, with named ratios and a named integration. The full mandate is in Part One → The mandate.

The friction, named

87% never make it through the door

That's not because the door is guarded. It's because the path to it is long, unclear, and public: a student has to know who to ask, be willing to be seen asking, and be willing to wait once they have. Every one of those is a friction point Sachetana is built to remove, ahead of a psychiatrist ever entering the picture.

It's also, separately, a market fact. Almost every serious competitor brings its own clinicians to earn, so none of them has any reason to reduce that friction by routing a student back to a university's own counsellors. MAHE has the opposite incentive, and the opposite asset: Kasturba Hospital, a Department of Psychiatry, and salaried counsellors it already pays for. No platform on the market is built to send a student there. Sachetana is.

Where sixteen platforms sit · full map in Part One

↔ Scroll to see the map

Depth of care
Student paysWho paysInstitution pays
BetterHelp
Amaha
Wysa
YourDOST
TimelyCare
Mantra Health
Sachetana
Brings its own clinicians Routes to the customer's own Where MAHE should sit
A low-friction, stigma-first front door, funded by the institution, that routes students into clinical resources the institution already owns, and gives that institution's clinicians the tooling to run it.

Finding 01

TimelyCare, Uwill, YourDOST and Talkspace all monetise the therapy hour; none has any incentive to hand a student back to a university's own psychiatrist.

Finding 02

No Indian platform ships the mandated 1:100 peer-supporter layer. At MAHE's scale, that's hundreds of people who need a system, not a spreadsheet.

Three more findings, the full sixteen-platform matrix, and the sourcing are in Part One → The gap.

How

The bridge, with side features layered on top

Sachetana does not replace MAHE's psychiatric care. It is the path to it. A student who would never call the counselling centre directly can start anonymously, get triaged, and be routed into the same clinicians and the same hospital MAHE already funds. Everything else the app does sits on top of that bridge, not instead of it.

Start here A student Notices something's wrong. Doesn't know who to ask, or whether it's "bad enough" to ask at all.
→↓
Sachetana The bridge Anonymous entry, AI triage, no visible line outside anyone's door.
→↓
Already exists MAHE's own care Kasturba Hospital · Dept. of Psychiatry · salaried counsellors.

Layered on top of the bridge, not instead of it

Mood tracking Journaling Safe Space peer support Resource library Follow-up nudges One-tap SOS → Tele-MANAS

The same bridge, broken into five tiers a student climbs

Tier 0Front door4 features · anonymous entry, AI triage
Tier 1Self-directed3 features · guides, mood tracking
Tier 2Community2 features · peer support, on- and offline
Tier 3Clinical2 features · real scheduling, psychiatry escalation
Tier 4Crisis1 feature · one-tap SOS, Tele-MANAS

Every card, its justification, and the phase it ships in are in Part Two → The ladder.

Go deeper

Four parts, read in order

Each part is built on the one before it. The competitor survey finds the gap; the feature set is that gap translated into a build; the business model is what it would cost MAHE to fund it; the compliance review is what legally attaches to the data any of it collects.

Part One

Competitor analysis
Covers
Sixteen platforms across five segments, a ten-capability feature matrix, six revenue models, and price anchors spanning four orders of magnitude.
Finding
Every commercial player must bring its own clinicians to earn. MAHE's incentive runs the opposite way, and nobody has built for that yet.
Read nextOpen Part One →

Part Two

Feature set
Covers
Fifteen features across five tiers of a stepped-care ladder, from an anonymous first message to a psychiatrist at Kasturba Hospital.
Finding
The 1:100 peer-supporter mandate is a named legal requirement that no Indian platform ships, the largest single gap in the entire market.
Read nextOpen Part Two →

Part Three

Business model
Covers
A recommended pricing shape, a phased build cost, a three-year run cost, and the return on a single utilisation number.
Finding
A phased build ballparked at ₹75 lakh–₹1.2 crore, recovered inside its first year or two purely from counselling utilisation MAHE already funds.
Read nextOpen Part Three →

Part Four

Regulatory & data
Covers
Six overlapping legal regimes, what each of the sixteen platforms actually collects where that's been independently checked, and what that means for this build.
Finding
A single unmade structuring decision determines whether the anonymous front door in Part Two is even legal for a seventeen-year-old.
Read nextOpen Part Four →

Prior art

We are not starting from a slide

Sachetana v1 was a working prototype, not a deployed product, built but never released to a live student population. It shipped self-evaluation, journaling, breathing practice, a Safe Space community feed, SOS support, counselling surfaces and a Gemini-based chatbot, on Flutter with Supabase and Firebase behind it, and covers the top of the funnel most competitors get wrong.

Self-evaluation Journaling Breathing practice Safe Space feed SOS support Counselling surfaces Gemini chatbot Flutter · Supabase · Firebase

Where v1 stops and the proposed feature set begins is mapped tier by tier in Part Two → Prior art.

Part one of four  ·  Competitor analysis

Sixteen platforms, five business models, one empty position.

A survey of who ships what, who pays for it, and where the market's money actually comes from, ending with the one position none of them occupy.

Mandate Market map Players Matrix Retention The gap

The mandate

The window is open, and it was opened by a court

Until last year, a campus mental health platform in India was a nice-to-have that competed with the library renovation for budget. That changed in July 2025. A Supreme Court judgment, followed by UGC action, converted student mental health from institutional goodwill into a compliance obligation with named ratios and named integrations.

This matters more than any market-size projection. MAHE now has to build a specific set of things. The question in the room is whether it buys them, assembles them from disconnected parts, or commissions them, and the mandate text reads remarkably like a product specification.

25 July 2025

Supreme Court, Sukdeb Saha v. State of Andhra Pradesh

Directed the Government of India to frame a uniform mental health policy for Higher Educational Institutions, built on the UMMEED guidelines, the Manodarpan initiative, and the National Suicide Prevention Strategy.

Mandated requirements

What every institution now owes its students

  • At least one qualified counsellor, psychologist or social worker at any institution with 100+ students
  • A mental health and wellbeing centre
  • Faculty mentors at 1:500
  • Peer supporters at 1:100
  • A 24×7 helpline integrated with Tele-MANAS
  • Staff training twice a year by certified professionals
13 November 2025

UGC advisory on Tele-MANAS

Higher Educational Institutions are asked to actively promote Tele-MANAS (14416) across campuses. The national programme has handled 29.82 lakh calls since launch, 12.33 lakh in 2025 alone, across 53 cells and 20 languages.

Read this as an integration requirement, not a competitor.

The peer-supporter ratio of 1:100 is, for a university of MAHE's size, a workforce of several hundred students who need recruiting, training, rostering, supervising and escalation paths. No product in the Indian market ships that. It is a software problem wearing a policy costume.

Market context, for completeness. India's mental health app market was valued at roughly $112.75M in 2023 and is projected at $363.46M by 2030 (18.2% CAGR). Global mental health technology: $15.22B in 2024 to a projected $30.98B by 2030. These numbers justify the category. They do not justify this project. The mandate does.

Market map

Where the money comes from, and how deep the care goes

Two axes explain almost every product decision in this market. Who pays determines whether the product must convert a distressed nineteen-year-old into a card-on-file, or simply has to be used. How deep the care goes separates a meditation library from a licensed psychiatrist.

The third variable is the one that does not fit on an axis, so it is drawn as a shape: whether a platform brings its own clinicians or routes into the customer's existing ones. Almost every player brings their own. That is the whole business.

Figure 1 · Positioning of sixteen platforms

↔ Scroll to see the full map

↑ Licensed clinical careInstitution-funded →
Depth of care
Student pays out of pocketWho paysInstitution or state pays
Calm
Headspace
Rocket Health
BetterHelp
Amaha
Talkspace
Lissun
Mpower
Wysa
SilverCloud
1to1help
YourDOST
Togetherall
YOU at College
Tele-MANAS
Uwill
TimelyCare
Mantra Health
Sachetana
Brings its own clinician network Routes into the customer's existing clinicians Our target position

Reading the map. The tinted region is not crowded because it is easy. It is empty because almost nobody can sell into it. A vendor with its own therapist network has no incentive to route a student to a university's psychiatrist; that is revenue walking out the door. MAHE has the inverse problem and the inverse incentive: Kasturba Hospital, a Department of Psychiatry established in 1974, a Department of Clinical Psychology, and salaried counsellors whose marginal cost per additional student session is already sunk. Positions are directional, compiled from public product pages and pricing, not from vendor disclosure.

The players

Sixteen platforms, five business models

Each entry below carries the same four facts: what they ship, who pays, how they make money, and the one thing we should take from them. Entries whose names are set in red are the ones that matter most to the MAHE conversation.

Global consumer wellness

D2C subscription · content-led

These built the category's habits and its vocabulary. They are not competitors for a university budget, but every student arriving at our app has been trained by them on what a wellness app feels like.

Calm homepage

Calm

US · Global
Ships
Guided meditation, Sleep Stories, breathwork, Daily Calm, mood check-ins, masterclasses, music, kids content.
Price
Premium $69.99–79.99/yr; Family $99.99/yr for six seats; lifetime $499.99. Calm Business for employers.
Model
D2C freemium subscription, with an enterprise arm selling seats to employers and health plans.
TakeContent alone sells beautifully and routes nobody to a clinician. Calm has no crisis pathway, no clinical governance, and no idea who its user is.
Headspace homepage

Headspace

US · Global
Ships
Meditation, mindfulness courses, sleepcasts, focus music, movement; coaching and therapy inside enterprise tiers post-Ginger merger.
Price
$12.99/mo · $69.99/yr consumer; enterprise priced per employee per month.
Model
D2C subscription plus enterprise and health-plan contracts.
TakeThe stack is converging on content, then coaching, then clinical. That sequencing is correct. Their unit economics assume an employer budget, which a campus does not have at the same per-head level.
Wysa homepage

Wysa

India-origin · Global
Ships
AI CBT conversational agent, 150+ self-help tools, DBT and mindfulness exercises, an SOS toolkit with grounding exercises, a personal safety plan and local helplines; human coach add-on; clinician-side Digital Referral Assistant.
Evidence
FDA Breakthrough Device designation, peer-reviewed RCT in JMIR, 6.5M+ users. NHS: 117,000+ patients through the referral assistant since 2022, saving clinicians ~21 minutes per assessment.
Price
Free anonymous tier; Premium ~$74.99/yr; coaching from $19.99/session; enterprise custom.
Model
B2B2C. Over 80% of revenue comes from businesses and health plans; the free tier is the funnel.
Closest analogue in the worldAnonymous-first AI front door, escalation to humans, and tooling that saves the clinician time. Built in Bengaluru. Their referral assistant is our thesis, proven, with a regulator's stamp on it.

Teletherapy marketplaces

D2C subscription → payer-funded

The cautionary segment. Both built enormous therapist networks on consumer credit cards, and both are now being rescued by institutional payers. Their trajectory is the single strongest argument for our B2B model.

BetterHelp homepage

BetterHelp

US · Teladoc
Ships
Matched licensed therapist, unlimited messaging, weekly live video or phone, journaling, group webinars.
Price
$70–100/week, billed four-weekly ($280–400/mo). Financial aid of 10–40% including a student discount.
Model
D2C subscription marketplace. Revenue fell 11% year-on-year in Q1 2025; acquired UpLift to move in-network with insurers.
TakeAt roughly ₹25,000 a month, this is structurally unavailable to an Indian undergraduate. Pure D2C therapy churns, and is now retreating toward third-party payers.
Talkspace homepage

Talkspace

US · NASDAQ: TALK
Ships
Messaging therapy, live video sessions, psychiatry, live workshops, teen and couples programmes.
Price
$69/wk messaging · $99/wk video and messaging · $109/wk with workshops. Roughly $10 average copay when insurance covers it.
Model
Deliberately shifted its core to an insured base. Payer business up 33% in Q1 2025, five consecutive profitable quarters, FY2025 guidance $220–235M.
The whole argument in one number$109 a week out of pocket versus a $10 copay. Demand did not change; the payer did. In our model, the university is the payer, which is precisely why the grant makes the product viable where a consumer app would die.

Campus-native platforms

B2B · per-student licence · our direct set

This is the segment MAHE would actually be choosing between. It is a real, funded, fast-growing market in the US, and almost entirely absent in India, which is both the opportunity and the reason no Indian procurement officer has a reference price.

TimelyCare homepage

TimelyCare

US · ~500 campuses
Ships
TalkNow 24/7 on-demand emotional support, scheduled counselling, psychiatry, health coaching, basic medical care, peer community, self-guided content, faculty referral tools. Acquired Alongside in May 2026 for clinician-powered AI coaching and risk detection.
Scale
~500 institutions, 3 million students. Inc. 5000 four years running on 254% revenue growth. $60M Series C.
Price
Three contract shapes: a 3% population cap, a 10% cap with unused visits rolling over, or all-students per-student-per-year. Public anchor: James Madison University, $435,600 in year one of a three-year deal.
Model
Institution pays; TimelyCare supplies the entire clinical network.
Benchmark and warningThe reference case for "a university buys a mental health layer." But it replaces the campus clinic rather than feeding it, and the whole apparatus is shaped around US insurance and state-by-state licensure. Neither constraint applies at MAHE, and neither does the price.
Uwill homepage

Uwill

US
Ships
Therapist choice and matching in minutes, video, phone, chat and message, 24/7 crisis line, wellness content.
Scale
$30M Series A from Education Growth Partners, $35.2M total. Inc. 5000 #79 in 2025 and #27 in 2024; fastest-growing company in Massachusetts, 2025.
Model
Institution-paid licence, typically hours-based or per-student.
TakeTheir entire pitch is time to first appointment, measured in minutes. Speed is the product. That is a metric we can beat with an in-house clinical team and no licensure friction.
Mantra Health homepage

Mantra Health

US · 52+ campuses
Ships
An explicit stepped-care ladder: self-care tools, a DBT-based life-skills programme, peer support via a Togetherall partnership, wellness coaching, teletherapy, telepsychiatry, and 24/7 crisis. Plus a Collaboration Portal giving campus clinicians visibility into referrals and student care plans.
Scale
~500,000 students across 52+ campuses including Penn State, MIT and Miami Dade. $22M Series A plus extension.
Model
Institution-paid, closed-loop referral between the campus counselling centre and Mantra's clinicians.
The structural blueprintThe Collaboration Portal is the single piece we must build for MAHE's Department of Psychiatry and its counsellors. Mantra bolts a campus onto its network; we would bolt a network onto the campus. Same architecture, inverted ownership; ours is the version a university actually wants.
Togetherall homepage

Togetherall

UK / US
Ships
24/7 anonymous, clinically moderated peer community, self-assessments, wellbeing courses on depression, sleep and anxiety, goal tracking, and signposting into each specific university's own resources.
Scale
Baylor, Boston University, Buffalo, Iowa, UW-Madison, UT El Paso, Pittsburgh.
Model
Institution-paid annual licence, sold explicitly as a cheaper tier that absorbs mild-to-moderate need so the counselling centre can concentrate on high-risk cases.
Validates the Safe SpaceProof that anonymous peer support is a legitimate, purchasable tier of campus care, and the evidence base for the feed Sachetana already shipped. The non-negotiable detail: clinically moderated, round the clock. An unmoderated student mental health forum is a liability, not a feature.
YOU at College homepage

YOU at College

US · Grit Digital Health
Ships
A personalised wellbeing portal across Succeed, Thrive and Matter domains, self-assessments, and a content engine whose main job is surfacing that campus's own resources.
Price
$0.75–$3.00 per student per year, varying by enrolment and customisation depth. 30+ campuses, 300,000+ students.
Model
Pure engagement and content SaaS, institution-paid, no clinicians involved.
The floor of the marketOur single most useful price anchor. A non-clinical engagement layer costs under $3 per student per year. Everything with clinicians in it costs roughly seven to twenty-five times that. Our proposal has to say clearly which of those two things it is.
SilverCloud by Amwell homepage

SilverCloud by Amwell

Ireland / UK / US
Ships
Guided internet-delivered CBT programmes with a human supporter model and built-in outcome measurement.
Model
Enterprise licence to health systems, employers and universities. Deployed across the NHS and HSE.
TakeSupported self-help, structured programmes with light human check-ins, is the most cost-efficient middle tier in the entire market. It converts one clinician's hour into many students' progress.

Indian clinical platforms

Freemium D2C + B2B hybrid

Where Indian capital and Indian clinical talent actually sit today. One of these is the incumbent MAHE would most plausibly buy instead of commissioning a build, so the comparison has to be made explicitly, not avoided.

YourDOST homepage

YourDOST

India · the incumbent
Ships
24/7 chat, audio and video with counsellors, anonymity-first onboarding, self-help content, wellness assessments, on-campus counsellor visits, workshops and webinars, a telephonic crisis helpline, and admin dashboards for the institution.
Campuses
IIT Bombay, IIT Kharagpur, IIT (ISM) Dhanbad, IIT Goa, NIT Rourkela, NIT Kurukshetra and others. NIT Kurukshetra's deal is explicitly a subscription giving students round-the-clock support free of charge.
Financials
FY24 revenue ₹19.46 Cr. Total funding $1.71M across three rounds: a capital-light business.
Model
Subscription B2B to campuses and corporates; free at the point of use for the student.
The question we must answer in the room"Why not just buy YourDOST?" The answer is ownership and marginal cost. YourDOST bills MAHE annually to supply its counsellors. MAHE already employs psychiatrists, clinical psychologists and counsellors, and runs Kasturba Hospital. Buying YourDOST means renting capacity the university already owns, forever, with the clinical relationship and all the data sitting outside the institution.
Amaha homepage

Amaha (ex-InnerHour)

India
Ships
500+ self-help tools, six structured courses covering depression, anxiety, sleep, stress, anger and happiness, mood tracking, guided breathwork and visualisation, sleep tools, small-group Mental Health Circles, a 50,000-member community, plus therapy and psychiatry online and at physical centres.
Scale
6M+ app users, 220k+ sessions delivered. FY24 revenue ₹22.76 Cr.
Price
Free self-help; therapy roughly ₹1,500–3,000+/session; psychiatry paid separately.
Model
Freemium D2C funnel converting into paid clinical services, with a B2B arm alongside.
TakeBest self-help library in India and the benchmark for content quality. Also the clearest illustration of the funnel problem: six million users, 220,000 sessions. On campus we delete the payment step entirely, which should move that conversion ratio by an order of magnitude.
Lissun homepage

Lissun

India
Ships
App plus physical centres, hospital partnerships, and a corporate arm; strong child and family focus.
Funding
$10.7M across 7 rounds from 87 investors; $2.5M pre-Series A led by RPSG Capital Ventures. Acquired US startup Being Cares Inc. in July 2025.
Model
Omnichannel: digital front end, physical clinics, institutional contracts.
TakeHybrid online and offline is where Indian capital is flowing, and it is the model MAHE is already structurally set up for: the app is the front door, the counselling centre is the room, Kasturba is the escalation.
Rocket Health homepage

Rocket Health

India
Ships
Online therapy and psychiatry; after a first consultation the user gets a personalised care roadmap and chooses single sessions or bundles.
Price
Sessions from ₹1,000, effectively the affordable end of private Indian therapy.
Model
Low-price D2C teletherapy, founded 2021.
TakeUseful as the market price of a therapy hour in India. It is the number to multiply against when valuing the sessions MAHE's own staff already deliver at zero marginal cost.
Mpower homepage

Mpower

India · Aditya Birla Education Trust
Ships
Centres of excellence, foundation clinics in Mumbai and Virar, a helpline, and school and college outreach programmes.
Model
Philanthropically backed clinical services and outreach. Founded 2016 by Neerja Birla; stated goal of impacting 150 million lives by 2030.
TakeThe trust-and-CSR funding route. Worth naming in the meeting as a plausible second funding source for scale-out beyond the initial university grant.
1to1help homepage

1to1help

India · EAP incumbent
Ships
Employee Assistance Programme: counselling, crisis response, wellbeing content, manager training, organisational reporting.
Scale
Founded 2001. 800+ corporates, 65+ cities, 2.7M employees covered.
Model
Annual contract priced per covered life, varying by organisation size, feature set and service level.
The pricing mechanic to borrowIndia already has a mature, twenty-four-year-old market for per-covered-life mental health contracts; it just sells to HR departments instead of registrars. Framing our proposal as an EAP for students gives MAHE's finance office a familiar, defensible procurement shape.

Public infrastructure

State-funded · free · mandated

The "why would we pay for this when it's free" objection, and the answer.

Tele-MANAS homepage

Tele-MANAS · 14416

Government of India
Ships
A two-tier service: trained counsellors provide first-level emotional support, psychosocial counselling and referral; psychiatrists and clinical psychologists are reachable at tier two through eSanjeevani. Free, 24×7, in 20 languages.
Scale
Launched 10 October 2022. 29.82 lakh calls handled as of November 2025, 12.33 lakh in 2025 alone, across 53 cells.
Model
Fully state-funded. Not a commercial competitor.
Turn the threat into a compliance featureUGC's November 2025 advisory asks institutions to actively promote Tele-MANAS. A helpline nobody knows the number to is not a service. Our SOS flow should dial 14416 directly, and our institutional reporting should evidence that promotion, which means the app becomes the mechanism by which MAHE demonstrates compliance rather than a duplicate of a free service.

Feature matrix

What everybody ships, side by side

Ten capabilities, sixteen platforms. The columns on the right, routing into the institution's own resources and giving that institution's clinicians a working surface, are where the market thins out dramatically.

Figure 2 · Capability coverage

↔ Scroll the table sideways

Compiled from public product pages, pricing pages and case studies. Directional, not vendor-verified.
Platform Anonymous entry Self-assessment AI triage / chat Self-help content Peer community Human counselling Psychiatry 24/7 crisis Routes to campus resources Clinician-side tooling
Calm ○○○●○○○○○○
Headspace ○◐◐●○◐◐○○○
Wysa ●●●●○◐○●○●
BetterHelp ○●○◐◐●○○○○
Talkspace ○●○◐◐●●○○◐
TimelyCare ○◐◐●◐●●●◐●
Uwill ○◐○●○●◐●◐●
Mantra Health ○●○●●●●●●●
Togetherall ●●○●●○○◐●◐
YOU at College ●●○●○○○○●◐
SilverCloud ○●○●○◐○○◐●
YourDOST ●●○●◐●◐●◐●
Amaha ◐●◐●●●●○○○
Lissun ○◐○◐○●●○○◐
Rocket Health ○◐○○○●●○○○
Tele-MANAS ●○○○○●●●○○
Sachetana v1 (2023) ●●●●●◐○◐◐○
● core, shipped ◐ partial, paid add-on, or via partner ○ not offered

What the last row says. Sachetana v1 already covered the top of the funnel: anonymous entry, self-evaluation, AI triage, content, peer feed, which is the half most products get wrong. What it never built is the bottom: real appointments with real MAHE clinicians, a psychiatry escalation path, and a working surface for the counsellors on the other side. That is the build.

Retention

The category's real failure is not access. It's the second week.

Every platform in this document can get a student to download. Almost none can get them to come back. If the brief is "engaging enough that students keep coming", this is the number that brief is really about, and it is worse than most people assume.

Figure 4 · Two views of the same drop-off

Median retention across 93 mental health apps

Share of installed users still active, by day

Day 1100%
Day 10≈ 20%
Over 80% of users are gone between day 1 and day 10
Day 153.9%
Day 303.3%
Best-in-class platforms reach roughly 16%

The friction funnel, per 100 students

Where motivated students are lost before they ever reach a counsellor

Enrolled students100
Report fair or poor mental health> 50
Use on-campus counselling13
Then wait two to eight weeks for the appointment
The addressable loss≈ 39 students
Students who need support and do not reach the resources their university already funds. This is the product.

What demonstrably works

  • Co-design with the actual students. Deep, participatory adaptation: language, cultural resonance, iterative refinement with the target population, is associated with dropout under 11% and adherence above 75%. That is a twentyfold swing against the category median.
  • Brevity and non-clinical language. In an Indian university co-design study, short flows and everyday words directly reduced stigma-related barriers to help-seeking.
  • A quick-exit button. Student-requested in that same study, and used by 90% of participants. A single control that lets someone leave the screen instantly, because being seen using the app is itself a barrier.
  • Clinical moderation of peer spaces, around the clock, as Togetherall does. It is what separates a support community from an incident.

What the data says to avoid

  • Shipping a content library and calling it a platform. Root-cause analysis of retention failure attributes roughly 40% to UX problems and 35% to inadequate content, meaning more content is not the fix.
  • Streaks and gamification borrowed from fitness apps. A broken streak is a failure notification delivered to someone already struggling.
  • Requiring identity at the front door. Personal stigma is significantly and negatively associated with help-seeking, and more so among male and younger students, precisely the cohort that dies.
  • Engagement as the success metric. Daily active users is the wrong goal. A student who opens the app twice and books a real appointment is a total success and a retention statistic.

The gap

The position nobody occupies

After sixteen platforms, the white space is specific and defensible. It is not "an Indian mental health app"; there are a dozen good ones. It is this:

A stigma-first front door, funded by the institution, that routes students into clinical resources the institution already owns, and gives that institution's clinicians the tooling and evidence to run it.
01

Every commercial player must bring its own clinicians

It is how they earn. TimelyCare, Uwill, YourDOST and Talkspace all monetise the therapy hour, so none of them has any incentive to hand a student back to a university psychiatrist. MAHE's incentive runs the opposite way; only a partner with no therapist network to protect can build honestly for that.

02

No Indian platform ships the mandated peer-supporter layer

The 1:100 peer supporter and 1:500 faculty mentor ratios require recruitment, training records, rostering, supervision and escalation: software nobody in this market sells. Togetherall proves anonymous peer support works; nobody has built the system that manages the supporters. At MAHE's scale that is hundreds of people needing a tool.

03

The clinician side of the product is almost universally neglected

Wysa's referral assistant saves NHS clinicians twenty-one minutes per assessment. Mantra's Collaboration Portal is its main institutional selling point. In India, the counsellor's experience is an afterthought, yet the counsellor is the scarce resource, and the person who decides whether the platform gets used at all.

04

Compliance evidence is unbuilt and now compulsory

After July 2025, MAHE must be able to demonstrate counsellor provision, Tele-MANAS promotion, peer supporter ratios and biannual staff training. Today that is a spreadsheet someone assembles under duress. A platform that generates it as a byproduct of normal operation is buying the administration something it cannot get anywhere else.

05

Nobody is optimising for the student who has already decided to get help

The category is obsessed with awareness and engagement. The brief here is narrower and more tractable: the student who is already motivated, already hurting, and stops at a form, a queue, a phone number, or the fear of being recognised in a waiting room. Every one of those is a solvable interface problem, and solving them requires no new clinical capacity at all.

Onward

What this fixes for Parts Two and Three

The feature set in Part Two follows directly from the five findings above; every proposed feature is tied back to one of them. The pricing and cost figures in Part Three are built on the six revenue models and the price anchors this analysis surfaced.

Part Two

Feature set
Open the stepped-care ladder →

Part Three

Business model
Open the pricing and cost case →

Part two of four  ·  Proposed feature set

A stepped-care ladder, not a feature list.

Fifteen features across five tiers, from an anonymous first message to a psychiatrist at Kasturba Hospital, each one justified by a mandate clause, a competitor precedent, or a retention finding from Part One, and each attached to the specific friction it removes.

Prior art Front door Self-directed Community Clinical Crisis Back office Rollout

Prior art

Where v1 stops and this ladder begins

Sachetana v1 was a working prototype, not a deployed product, built but never released to a live student population. It shipped self-evaluation, journaling, breathing practice, a Safe Space community feed, SOS support, counselling surfaces and a Gemini-based chatbot, on Flutter with Supabase and Firebase behind it.

Against the sixteen platforms surveyed in Part One, that covers the top of the funnel: the half of the product most competitors get wrong, and the half that costs the most to get right. What follows is everything between that and a full stepped-care ladder: the clinical bottom, the community layer the mandate requires, and the institutional back office.

Shipped in v1, extended below, not rebuilt

Self-evaluation Journaling Breathing practice Safe Space feed SOS support Counselling surfaces Gemini chatbot Flutter · Supabase · Firebase

The gap Part One identified, closed tier by tier below

Real booking into MAHE clinicians Counsellor-side working surface Psychiatry escalation path Clinically moderated peer layer Peer-supporter management (1:100) Tele-MANAS integration Compliance reporting Quick exit

The ladder

Fifteen features, five tiers, in the order a student climbs them

Nobody starts at the psychiatrist. Every feature below sits at the point on that path where a competitor, a court order, or a piece of retention data says something specific has to happen, and each card names which.

Tier 0: The front door

4 features · Phase 1

Everything a student sees before they have decided to trust anyone with their name.

Anonymous-first entry

Phase 1
What
No login wall, no name, no student ID required to open the app or start a conversation. Identity is only asked for at the point an appointment must actually be booked.
Removes
The fear of being on record as someone who needed help, before any help has even been offered.
Retention findingPersonal stigma is significantly and negatively associated with help-seeking, and more so among male and younger students, precisely the cohort at highest risk. Togetherall and YourDOST both lead with anonymity for the same reason.

AI triage chat

Phase 1 · extends v1
What
A conversational agent, the Gemini chatbot already in v1 given structured triage logic, that listens, gauges severity, and either resolves the moment or routes to the right human tier.
Removes
Not knowing whether what you're feeling is "bad enough" to justify asking a real person for help.
Wysa precedentWysa's clinician-facing referral assistant saves NHS staff roughly 21 minutes per assessment and carries an FDA Breakthrough Device designation. Same idea: let the AI do the sorting so a scarce human only sees what needs a human.

Validated self-assessment

Phase 1 · extends v1
What
Short, private screening tools that give the triage layer a structured signal instead of a guess, building on the self-evaluation already in v1.
Removes
Not having language for what's wrong; the assessment supplies the vocabulary.
Category benchmarkShipped by nearly every serious platform in Part One: Amaha, Togetherall, TimelyCare, YourDOST. Its absence would be conspicuous, and it also makes the AI triage layer more accurate.

Quick-exit control

Phase 1 · new, small
What
One control, always visible, that instantly replaces the screen with something innocuous.
Removes
Being seen using the app: in a hostel room, a library, a lecture hall.
Retention findingStudent-requested in an Indian university co-design study, and used by 90% of participants once it shipped. Small to build, disproportionately used.

Tier 1: Self-directed care

3 features · Phase 1

Help that doesn't consume a counsellor's time, and is not allowed to pretend it replaces one.

Wellbeing guide library

Phase 1 · extends v1
What
Structured, short-form content on sleep, exam stress, anxiety and low mood, organised as guided sequences, not an undifferentiated blog.
Removes
The step between "I feel bad" and "I know what might help", for the large majority of students whose need is real but not clinical.
Amaha precedent500+ self-help tools is the category's content benchmark. Content is a fixed cost served to everyone; every student it satisfies is a session it did not have to take from a counsellor's calendar.

Mood tracking & journaling

Phase 1 · extends v1
What
The lightweight daily check-in and journal already in v1, kept deliberately simple rather than expanded into a wellness-tracking suite.
Removes
Nothing on its own: it's a habit loop, not a fix. It exists to bring a student back on an ordinary day, before a bad one.
Retention findingCategory-wide churn analysis attributes ~35% to inadequate content and ~40% to UX. This only earns its place paired with the design discipline at right; alone, it's exactly the mistake that finding warns against.

Non-gamified engagement

Principle, applies everywhere
What
No streaks, no leaderboards, no badge count. Success is measured as task completion: a session booked, a course finished, never as daily-active-use.
Removes
Guilt, manufactured urgency, and the specific harm of a broken streak.
Retention finding"A broken streak is a failure notification delivered to someone already struggling." Fitness-app mechanics are the most-cited anti-pattern in the category's own retention research.

Tier 2: Community care

2 features · Phase 1 → Phase 2

Help from other people, before help from a clinician: one layer already exists, one has to be built.

Online peer support (Safe Space v2)

Phase 1 · extends v1
What
An anonymous, clinically moderated community feed, around the clock, extending the partial version already shipped in v1's Safe Space.
Removes
The isolation of believing no one else on campus is going through the same thing.
Togetherall precedentTogetherall's whole business proves anonymous peer support is a legitimate, purchasable tier of care. The non-negotiable detail borrowed directly from them: moderation must be clinical and continuous, or the feature becomes a liability.

Offline peer-supporter network

Phase 2 · new build
What
Recruitment, training records, shift rostering, supervision and escalation tooling for a peer-supporter workforce drawn from MAHE's own students, plus tracking for faculty mentors.
Removes
For MAHE specifically: converts an unfunded compliance obligation into a managed programme, instead of a spreadsheet nobody owns.
Mandate clauseThe Supreme Court order names ratios of 1:100 peer supporters and 1:500 faculty mentors. Part One's survey found no Indian platform shipping this, the largest single gap identified in this market.

Tier 3: Clinical care

2 features · Phase 2 → Phase 3

MAHE's own psychiatrists and counsellors, with the queue and the guesswork removed.

Real session scheduling

Phase 2 · new build
What
Live booking into MAHE's actual counsellors: calendar and availability visibility, and a waitlist position instead of silence, replacing the "counselling surface" v1 shipped without real appointments behind it.
Removes
The two-to-eight-week wait a student hits after finally deciding to ask, and not knowing whether the request even reached anyone.
TimelyCare precedentThis is the single largest gap Part One identified. TimelyCare and Mantra Health both treat scheduled counselling as core infrastructure; v1 got everything before this point right and stopped exactly here.

Psychiatry escalation path

Phase 3 · new, integration
What
A defined clinical handoff from counselling into MAHE's Department of Psychiatry and Kasturba Hospital for cases beyond a counsellor's scope, carrying context forward instead of restarting the story.
Removes
Re-explaining everything to a new professional, and the drop-off between "this needs a psychiatrist" and an actual appointment existing.
Mantra Health precedentMirrors Mantra's explicit stepped-care ladder, the structural blueprint Part One identified, inverted here so MAHE owns every clinician in the chain instead of renting them.

Tier 4: Crisis

1 feature · Phase 1

When it can't wait for an appointment of any kind.

One-tap SOS with Tele-MANAS dial-out

Phase 1 · extends v1
What
A single always-reachable control that surfaces the nearest emergency resource, a saved personal safety plan, and a direct dial to Tele-MANAS's 14416, building on the SOS support already in v1.
Removes
Having to search for a number, or decide which service to call, in the moment it matters least to be searching.
UGC advisoryThe November 2025 advisory specifically directs institutions to actively promote Tele-MANAS. Wiring the dial-out into the app both delivers on that and gives MAHE evidence that it complied.

Back office

Built for MAHE's staff, not just its students

Every feature above only works if the people on the other end of it aren't fighting the product too. This is the half of the market Part One found almost universally neglected.

Cross-cutting

3 features · Phase 2 → Phase 3

Counsellor-side working surface

Phase 2 · new build
What
A caseload view for MAHE's counsellors and psychiatrists: session notes, referral flags surfaced by the AI triage layer, and visibility into the peer-supporter roster.
Removes
The counsellor's side of every piece of friction above. A front door nobody staffs is not a front door.
Mantra Health precedentWysa's version saves NHS clinicians 21 minutes per assessment; Mantra's Collaboration Portal is its main institutional selling point. In India, this side of the product is close to an afterthought everywhere.

Compliance & reporting dashboard

Phase 3 · new build
What
Auto-generated evidence of counsellor-to-student ratio, peer-supporter ratio, Tele-MANAS promotion activity, and biannual staff training, assembled as a byproduct of normal use.
Removes
The spreadsheet someone at MAHE currently has to build under duress whenever this compliance is questioned.
Mandate clauseAfter July 2025 this evidence is compulsory, not optional. A platform that produces it automatically is selling the administration something no competitor in Part One offers at all.

Multilingual interface

Phase 3 · new, localisation
What
Hindi and regional-language interfaces alongside English, matching the twenty languages Tele-MANAS itself already operates in.
Removes
Excludes exactly the students least likely to already know how to ask for help in a second language, from a product whose entire purpose is reducing that barrier.
Category benchmarkTele-MANAS ships this at national scale; MAHE draws students from across India and abroad. A stigma-reduction product that only speaks English is a contradiction of its own premise.

Rollout

What ships when

Sequenced so MAHE commits to each stage against demonstrated use of the one before it, not all fifteen features in a single sitting.

What each phase costs to build, and what it costs to run afterward, is in Part Three → Build and run cost.

Part three of four  ·  Business and revenue model

What MAHE would actually be signing up to pay.

A recommended contract shape, a ballpark build cost, and the return on a single number: how many more students the university's own counsellors could see.

Revenue models Price anchors Our pricing Build & run cost ROI case The ask

Revenue models

Six ways this market makes money

Every platform above runs one of these six, or a blend of two. Only three survive contact with an Indian university campus, and the reasons the other three fail are worth stating out loud before anyone in the meeting proposes them.

Consumer subscription

Calm · Headspace · Wysa Premium

A free tier builds the habit; $70–80 a year unlocks the library. Margins are excellent because content is a fixed cost served infinitely.

Wrong for us

Asking a student in distress for a card is the exact friction we exist to remove. It also re-imports stigma: paying marks you as someone with a problem.

Pay-per-session marketplace

BetterHelp · Talkspace · Amaha · Rocket Health

Take a margin on each therapy hour. Scales with supply of clinicians, and customer acquisition cost is brutal because every user churns after recovery.

Wrong for us

MAHE's clinicians are salaried. There is no margin to take on a session the university already paid for, and charging students for access to their own faculty would be indefensible.

Per-student annual licence

YOU at College · TimelyCare · Togetherall

The institution pays an annual fee multiplied by enrolment. Revenue is predictable, procurement understands it, and the student never sees a price.

Primary model

This is the shape the finance office already knows how to approve. It maps cleanly onto MAHE's existing per-student fee structure and renews annually with the academic year.

Capped-utilisation contract

TimelyCare's 3% and 10% population caps

The institution buys a pool of sessions sized to a percentage of enrolment rather than the whole population. Unused visits may roll over.

Useful hedge

Worth proposing as a second-year option if MAHE balks at an all-students figure. It also gives us a defensible way to price surge capacity during exams without owning idle clinician time.

Per-covered-life EAP

1to1help · corporate wellbeing contracts

An annual contract per covered person, tiered by feature set and service level, with organisational reporting as a core deliverable rather than an extra.

Framing to borrow

Twenty-four years of Indian precedent for exactly this contract shape. Presenting Sachetana as a student assistance programme rather than "an app" changes which budget line it competes for.

Grant, CSR and philanthropy

Mpower · Tele-MANAS · university innovation grants

Capital comes from a trust, a CSR budget or the state. No revenue requirement, but also no automatic renewal: funding must be re-justified with outcomes.

How we start

The university grant funds the build. But grants fund projects, not products, so the proposal must name the recurring model that takes over at the end of it, or the platform quietly dies in year two.

Price anchors

What this actually costs, per student, per year

Nobody in the meeting will have a reference price, because the campus mental health category barely exists in India. These are the anchors worth putting on the table, on a logarithmic scale, because the range spans four orders of magnitude.

Figure 3 · Annual cost per person, USD, log scale

Institutional licences in black · our reference band in red · individual retail in grey

↔ Scroll the scale sideways

$1
$10
$100
$1,000
$0.75 – $3.00
YOU at College: engagement layer, no clinicians
₹300 – ₹1,200
Indian campus SaaS norm, per student per year
≈ $21
TimelyCare: derived from JMU's $435,600 year one
$70 – $80
Calm / Headspace / Wysa Premium: individual
$215 – $430
Amaha: 12 therapy sessions at Indian retail
$3,588 – $5,668
Talkspace / BetterHelp: US retail teletherapy

The gap we have to price into

Between $3 and $21 per student per year sits the entire difference between a content portal and a clinical service. TimelyCare charges roughly seven to twenty-five times what YOU at College does, because TimelyCare employs the therapists.

We employ none. MAHE does. That is what lets a proposal land credibly below the clinical band while delivering clinical outcomes: the expensive input is already on the university's payroll.

Two numbers for the grant conversation

A therapy hour in the Indian private market costs a student ₹1,000–₹3,000. Every session MAHE's own counsellors deliver through the platform is that value returned to a student at zero marginal cost.

If the platform moves campus counselling utilisation from the benchmark 13% toward even 25% at a university of MAHE's scale, the delivered value is calculable in crores, and it is the cleanest ROI argument available for Part Two.

Method note. The TimelyCare figure is inferred by dividing a publicly reported contract value by approximate enrolment; treat it as an order-of-magnitude anchor, not a quoted rate. Rupee conversions use approximately ₹83 to the dollar. Vendors in this category do not publish institutional pricing, so every institutional figure here is either publicly disclosed by a customer or derived.

Our pricing

What we would actually ask MAHE to pay, per student, per year

The floor is YOU at College's $3 content-only price. The ceiling is TimelyCare's $21, and that number belongs to a vendor supplying its own therapists. We supply none: MAHE's psychiatrists and counsellors are already on payroll. Our price is for the software and workforce-orchestration layer sitting on top of clinicians MAHE already owns, so it should read as a fraction of the clinical band, because it is doing a fraction of the clinical work.

Figure 5 · Proposed price by phase, illustrative

↔ Scroll the table sideways

Enrolment figures are illustrative; MAHE's own headcount by campus should replace them before any figure reaches finance.
Phase What's live Per student / year Manipal campus (~26,000) All MAHE campuses (~40,000+)
Phase 1 Front door, self-directed care, online peer support ₹165–330 ($2–4) ₹43L–86L ₹66L–1.3Cr
Phase 2 + offline peer network, real scheduling, counsellor dashboard ₹415–750 ($5–9) ₹1.08–1.95Cr ₹1.66–3.0Cr
Phase 3 + psychiatry escalation, compliance reporting, multilingual: steady state ₹665–1,000 ($8–12) ₹1.73–2.6Cr ₹2.66–4.0Cr

Method note. MAHE's own admissions materials state 40,000+ students across all campuses; the Manipal campus, where Kasturba Hospital and the Department of Psychiatry sit, is the more relevant base case for phases that route into a specific clinical workforce. Rupee figures use ≈₹83 to the dollar, consistent with Part One.

Why this isn't a subscription number

This is proposed as an institutional line item: an annual contract sized to enrolment, in the shape 1to1help already sells to Indian corporates as an Employee Assistance Programme. The student never sees a price, exactly as in the "per-student annual licence" model Part One identified as the primary fit.

The capped-utilisation fallback

If MAHE prefers not to commit to an all-students figure in year one, borrow TimelyCare's cap structure: price a pool sized to 15–20% of enrolment, with unused capacity rolling over. A lower initial ask, with a defined trigger to expand.

What the grant needs to fund

A ballpark build cost, phased so MAHE isn't asked to commit to all of it at once

Three phases, matching the rollout in Part Two. Costs are fully-loaded Indian team estimates, not quotes, and should be treated the same way as every other order-of-magnitude figure in this dossier.

Phase 1 · 9–12 months

₹35–55 lakh

Two full-stack engineers, one product/UX designer, a part-time clinical content lead, and founder/PM time. The bulk of this phase extends an existing codebase rather than starting one.

Phase 2 · 6–9 months

₹25–40 lakh

Fewer people, harder problems: the scheduling system, MAHE calendar and systems integration, and the counsellor-side dashboard. This is where most of the engineering risk actually sits.

Phase 3 · 4–6 months

₹15–25 lakh

Psychiatry-workflow integration, the compliance dashboard, and localisation into Hindi and regional languages.

Total: roughly ₹75 lakh to ₹1.2 crore across nineteen to twenty-seven months, phased so each stage is committed against demonstrated use of the one before it, not promised in a single sitting.

What it costs to keep running

  • Hosting and infrastructure at 26,000–40,000-user scale: ₹4–8 lakh/yr
  • A lean maintenance team, one to two engineers: ₹15–25 lakh/yr
  • Clinical content review and updates: ₹4–6 lakh/yr
  • Peer-supporter programme coordination, a role that can sit inside student affairs: ₹3–5 lakh/yr

Total: roughly ₹26–44 lakh a year at steady state.

Why that number is comfortable

Even the lowest Phase 3 licence estimate for the Manipal campus alone, ₹1.73 crore a year, is roughly four to six times this run cost. The platform does not need to break even against a subscription; it needs to beat its own maintenance bill, and at these figures it does so with wide margin, whether that headroom is booked as an internal budget line or eventually offered to sister institutions.

The return, in one number

The argument is a utilisation curve, not a revenue projection

Every session MAHE's own counsellors deliver through this platform is value the university already funds, reaching a student who currently isn't reached. That is a cleaner case for a grant committee than any subscription math.

Figure 6 · Campus counselling utilisation, current vs. achievable

Current utilisation13%
≈ 3,380 of Manipal's ~26,000 students, at the College Pulse benchmark
Achievable utilisation25%
≈ 6,500 students: the ceiling reached in culturally co-designed programmes, per Part One's retention findings
≈3,120 additional students seen a year, at a private-market value of ₹1,000–3,000 a session, is ₹31 lakh to ₹94 lakh in equivalent value delivered annually, from a utilisation increase alone, without MAHE hiring a single additional clinician.

Method note. Session value is anchored to Rocket Health and Amaha's Indian retail rates, cited in Part One. The delta is 25% minus 13% of ~26,000 Manipal students; this is illustrative order-of-magnitude reasoning, not a forecast, and should be rebuilt on MAHE's actual current utilisation once that figure is available internally.

The ask

What to put on the table

₹75L–1.2Cr
phased build cost across three stages, roughly 19–27 months
Part Three · build cost
₹26–44L/yr
to keep it running at steady state, once built
Part Three · run cost
₹31–94L/yr
in equivalent value from a utilisation lift alone, before counting anything else
Part Three · ROI case
13%→25%
the single number this entire proposal is trying to move
Utilisation, current → achievable

None of these figures should survive contact with MAHE's actual enrolment numbers and department budgets unchanged. What should survive is the shape of the argument: a grant funds the build, a modest per-student line or an internal budget allocation funds the running of it, and the return is measured in students who reach a clinician MAHE already employs, not in downloads.

Part four of four  ·  Regulatory and data compliance

What this app is legally allowed to know about a student.

Six overlapping regimes govern the data a stigma-first mental health platform would collect, from a Supreme Court mandate already covered in Part One to a data-protection law still phasing in through 2027. Six of the sixteen platforms surveyed there have had their data practices independently checked, by researchers or by regulators; the results range from a mechanism worth copying outright to a $7.8 million federal fine. This part maps what's legally collectible, what the sixteen actually collect, and what MAHE's own build should and shouldn't ever retain.

The landscape The minors problem What Sachetana would generate What the sixteen collect When it goes wrong Design implications

The landscape

Six regimes, none of them written with an app like this in mind

No law on this list was drafted for a stigma-first triage chat that talks to a student before anyone knows their name. Each was written for something adjacent, a hospital, a bank, a European data controller, and each still reaches this product anyway. What follows is what actually governs Sachetana's data, not what a generic privacy-policy template assumes does.

DPDP Act, 2023 + Rules, 2025

India · phasing in to 2027
Governs
All personal data processed digitally, health data included, with no separate "sensitive data" category, a deliberate departure from both the EU and India's own 2011 rules.
The catch
Anyone under 18 needs verifiable parental consent and cannot be profiled or shown targeted content, unless the processing is done by a clinical establishment or mental health institution, an exemption written into the rules themselves.
Why it matters hereA meaningful share of MAHE's first-year cohort is 17 or 18. Whether Sachetana keeps its anonymous, no-login front door for them depends entirely on how the platform is structured, not on how it's used.

IT Rules, 2011 (SPDI Rules)

India · still the working baseline
Governs
The pre-DPDP definition of sensitive personal data, still the operative standard until the DPDP Rules' consent and security provisions fully commence. Rule 3 names physical, physiological and mental health condition, and medical records and history, explicitly.
The catch
It survives as India's only functioning cross-reference for "is mental health data sensitive," even though the law that supersedes it declines to use the word at all.
Practical readingTreat Sachetana's data as sensitive under the old rule and the new one at once. The obligation doesn't disappear because DPDP stopped naming it.

Mental Healthcare Act, 2017 §23

India · sector-specific, absolute by default
Governs
A right to confidentiality for anyone receiving mental healthcare, extending explicitly to information stored in electronic or digital format, in real or virtual space.
The catch
Disclosure is permitted only to a nominated representative, or to prevent harm and only on the order of a High Court or the Supreme Court. There is no lower bar, no institutional exception, no research carve-out.
Where it bitesThe moment Sachetana's AI triage or a counsellor's notes constitute "mental healthcare," Section 23 outranks every consent checkbox in the product.

ABDM Health Data Management Policy

India · optional infrastructure
Governs
The consent-manager framework behind Ayushman Bharat Health Accounts: explicit, purpose-specific consent, and a right to be forgotten with Health ID deletion.
The catch
Linking Sachetana into ABDM is a choice, not an obligation, and it adds a second consent layer on top of DPDP and the Mental Healthcare Act rather than replacing either.
RecommendationDon't link mental health records into ABDM by default. The interoperability benefit is real; the exposure it buys, for the single most sensitive data category this platform has, usually isn't worth it yet.

Global comparators the sixteen operate under

US · EU · for context
Governs
HIPAA (US clinical relationships only), GDPR Article 9 (EU, health data as a special category needing explicit consent), FERPA (US campus counselling records, not HIPAA, per joint federal guidance).
The catch
A consumer wellness app can sit entirely outside HIPAA by never becoming a "covered entity." Several of the sixteen do exactly that, legally, and it shows in what they collect.
Read nextWhat that gap looks like in practice is the next section but two.

The Supreme Court order and the counsellor-ratio mandate these regimes sit alongside are covered in full in Part One → The mandate; this page is about the data, not the staffing.

What Sachetana would generate

Every proposed feature is also a data-collection decision

Part Two justifies each feature by the friction it removes. The same fifteen features also define exactly what health data would exist, in whose name, and under which of the four Indian regimes above. Read side by side, the two lists are the same list.

Figure 7 · What each tier of the ladder actually collects

↔ Scroll the table sideways

Legal treatment is this dossier's reading, not a legal opinion; MAHE's own counsel should confirm before any of this ships.
Generated by What exists Governed primarily by Recommended posture
AI triage chat · Tier 0 Free-text conversation, pre-identity DPDP as ordinary personal data; MHCA in substance if content discloses a mental health condition Redact identifiers on detection; never store a name alongside content
Validated self-assessment · Tier 0 Screening scores, pre-identity DPDP; functions as a diagnostic signal regardless of what it's labelled Keep unlinked to identity until a human counsellor is looped in
Mood tracking & journaling · Tier 1 Longitudinal self-reported data DPDP; the closest analogue to a personal health record outside a clinical visit Student-owned: exportable and deletable on demand
Peer support · Tier 2 Posts, semi-public within the platform DPDP; MHCA confidentiality risk if identity is inferable in a small cohort Anonymity guarantees at least as strong as Togetherall's
Session scheduling · Tier 3 Real identity attached to a mental health record for the first time Mental Healthcare Act, Section 23, in full Never leaves MAHE's own systems; no processor, no exception
One-tap SOS · Tier 4 Crisis event, timestamp, Tele-MANAS handoff MHCA; the exact dataset a 2022 US case study shows gets mishandled first No secondary use of any kind, written into the vendor agreement itself

Row four's Tier 3 boundary is the one that matters most: everything above it can be built exactly as Part Two describes without triggering full MHCA confidentiality, and everything at or below it can't be built any other way.

What the sixteen collect

Six have been checked. The other ten haven't.

Nobody publishes a "data collected" column next to their pricing page, so this section leans on the handful of independent reviews that exist rather than vendor privacy policies, which say what a company intends, not what its software does. Figure 8 orders what's actually been found, best practice to worst.

Figure 8 · Independently verified data practices, best to worst

Wysa

India-origin · the benchmark
Found
Highest score, 6 of 7, in SFLC.in's September 2025 review of five apps against India's DPDPA principles; one of only two apps to earn Mozilla Foundation's Privacy Not Included "Best Of" citation in its 2022 review of thirty-two mental health and prayer apps.
Mechanism
No personal identifiers are required to use the core product. Identifiers a user volunteers inside a chat are, by the company's own account, redacted from its database within 24 hours of detection.
Worth copyingRedaction-by-design is a mechanism, not a policy promise. It's the single most directly reusable idea in this entire section.

BetterHelp

US · Teladoc
Found
Fined $7.8 million by the US FTC in March 2023 for sharing users' health information, including answers to intake questions about mental health history, with Facebook, Snapchat, Criteo and Pinterest for advertising, after telling users it wouldn't.
Mechanism
Advertising and analytics pixels embedded in the platform passed sensitive data to ad networks as a side effect of ordinary marketing tooling, not a deliberate data sale.
Why it matters hereThe mechanism, not the intent, is the lesson: ad and analytics SDKs are the most common way sensitive health data leaves a platform without anyone deciding to sell it.

Talkspace

US · NASDAQ: TALK
Found
Scored lowest, 1 of 7, in the same SFLC.in review, and named among the worst three apps, alongside BetterHelp, in Mozilla's 2022 review. Former employees told the New York Times that anonymised therapy transcripts were routinely mined for marketing insight.
Disputed
Talkspace denies the specific claim and states the platform is HIPAA, HITECH and SOC2 audited.
Contested, not provenWhat's independently verifiable is the low third-party data-protection score, from a different methodology entirely, arriving at the same conclusion.

Headspace & Calm

US · global consumer wellness
Found
Both scored 2 of 7 in the SFLC.in review, driven mainly by data-minimisation and consent-transparency failures rather than any single incident. Calm added full account-level data access and deletion for all users in June 2022, after Mozilla's review.
Pattern
No scandal, no fine, just routine over-collection.
The quiet failure modeDemographic inference and behavioural analytics, layered onto a meditation app because the SDKs that provide them are the industry default, not because anyone needed the data.

YourDOST

India · the incumbent
Found
Also scored 2 of 7 in the SFLC.in review: the same band as Calm and Headspace, and the only Indian platform of the sixteen with an independent, published data-protection assessment located for this dossier.
The gap this creates
No comparable published review exists for Amaha, Rocket Health, Lissun, Mpower, 1to1help, Uwill, Mantra Health, Togetherall, YOU at College, SilverCloud or Tele-MANAS.
The actual findingThat absence is itself the finding: nobody is checking, and MAHE would be the first institution in this market to ask.

Method note. Scores are from SFLC.in's review of five apps against DPDPA principles, published 29 September 2025; the Mozilla Foundation's 2022 review of thirty-two mental health and prayer apps; and the US FTC's March 2023 settlement with BetterHelp. Directional, not vendor-verified, in the same sense as Part One's capability matrix.

When it goes wrong

The pattern outside MAHE's sixteen looks the same

None of the three cases below are platforms in Part One's survey. They're here because each involved the same category of data Sachetana's crisis layer would generate, and each shows a different way that data got out.

01

Cerebral, $7 million+, April 2024

The FTC and DOJ found the telehealth prescriber shared the health data of 3.2 million patients, including psychiatric history and medication records, with TikTok, Meta, Google, LinkedIn and Snapchat for ad targeting, while its privacy policy said otherwise.

02

GoodRx, $1.5 million, February 2023

The first enforcement action ever brought under the FTC's Health Breach Notification Rule, for the same category of advertising-tracker leak. It set the precedent BetterHelp was fined under a month later.

03

Crisis Text Line, January 2022

The operator of a suicide-prevention text line shared texters' conversation data with Loris.ai, a for-profit spinoff building customer-service chatbots. A US FCC commissioner's public letter and a Politico investigation ended the arrangement within three days, but the data had already moved.

Every one of these three sits closer to Sachetana's own crisis tier than to a meditation app: a person in acute distress, a chat or text transcript, and a vendor relationship signed by someone who never anticipated the data being repurposed. The one-tap SOS integration with Tele-MANAS in Part Two → Crisis should carry an explicit no-secondary-use clause in whatever agreement sits behind it, not a generic vendor NDA.

Design implications

What this changes about the build in Part Two

None of this is a reason to slow down. It's a short list of decisions that are nearly free to make now and expensive to unwind once the platform has students' names attached to the mental health records those students trusted it with before it asked for one.

Build this in from day one

  • Structure Sachetana as an extension of Kasturba Hospital's clinical establishment, not a stand-alone app it refers into. That's what plausibly earns the DPDP Rules' children's-consent exemption for a large share of MAHE's cohort.
  • Redact identifiers from free text on detection, the way Wysa does, rather than relying on the anonymous-first design alone. Anonymity at the front door doesn't protect content a student volunteers inside it.
  • Treat the moment identity meets a mental health record, real session scheduling in Tier 3, as the one hard boundary in the system. Section 23 of the Mental Healthcare Act applies in full from that point on, no exceptions.
  • Write no-secondary-use into the Tele-MANAS SOS integration explicitly, rather than trusting a standard vendor agreement to cover it.

What the evidence says to avoid

  • Default analytics and advertising SDKs. BetterHelp's $7.8 million fine came from ordinary ad-tracking pixels, not a decision to sell data.
  • Linking mental health records into ABDM by default. The interoperability is real; for this specific data category, the added consent-manager exposure usually isn't worth it yet.
  • Implied consent by continued use. SFLC.in's review found this the most common failure among the apps it checked, including two of the four global names in Part One's matrix.
  • Any processor for crisis-tier data without a written no-reuse clause. Crisis Text Line's data reached a for-profit spinoff through exactly this gap.

Every recommendation above is cheaper to implement in the design phase Part Two describes than to retrofit afterward. None of them changes what the platform does for a student; all of them change what happens to the record of it.