जननी सुरक्षा
AI-Powered Maternal Health Risk Intelligence
Predicting which pregnancies will turn dangerous — before it's too late.
The intelligence layer India's ASHA network was built for — but never got.
An initiative of dmj.one
Vision: Aatmnirbhar Viksit Bharat 2047
arrow keys / click to navigate • F = fullscreen
The Problem
India built the world's largest community health worker network. Then gave them paper and a pen.
~19,000
deaths per year (UN MMEIG 2023, MMR 88)
1.18M
ASHA workers visit every pregnant woman monthly
0
have an AI tool to predict risk
Zero. Not one.
Most of these deaths are preventable, as the health-care solutions to prevent or manage complications are well known — if detected and reached in time. (WHO)
Value Proposition
From passive registration to active prediction.
We don't sell software. We sell time — the minutes between a complication and a death.
O(1) constant-time. No internet needed. Works offline.
Routes to facilities that can ACTUALLY help. Precomputed shortest paths.
Predicts crisis weeks before it happens. Trajectory-based forecasting.
5 minutes per assessment. Zero typing. Voice-first in 12 languages.
Underlying Magic
Addressing a capability gap — no comparable predictive tool currently exists for ASHA workers.
70K
Bayesian risk entries — single hash lookup. Beta-Binomial conjugate priors.
10,602
Real geocoded facilities from data.gov.in National Hospital Directory. Precomputed spatial index + capability matching.
7,480
Hb trajectory profiles — learned index prediction. IFA compliance modeling.
~50MB
Total memory
No GPU
No database queries
5ms
Response time
Offline
Works everywhere
Literature-informed: All 12 risk factors cross-validated against 5 independent data sources (NFHS-5, WHO, Cochrane, Lancet, ACOG). 12/12 within published confidence intervals.
Business Model
50 rupees per pregnancy. 27 million pregnancies. Do the math.
= Rs 135 crore TAM
Per-ASHA or per-pregnancy licensing to state National Health Mission departments.
Real-time monitoring dashboards for District Health Officers and CMOs.
MCTS/RCH portal integration, ambulance system connectivity (108/102).
Current
Free demo on GCP
Scale-to-zero = ~$0/month
Global
80+ countries via DHIS2
Adaptable architecture
Go-to-Market Plan
We don't need to convince 1.18 million ASHAs. We need to convince 1 District Health Officer.
A high-MMR district — 50 ASHAs, 1,500 pregnancies. Prove concept with measurable outcomes.
500+ ASHAs, cluster-randomized controlled trial to validate literature-informed risk model in the field.
Uttar Pradesh or Assam — state-wide deployment through NHM integration.
Integration with national health infrastructure. 27 million pregnancies per year.
Entry point: District Health Officer — they control budgets and ASHA deployment.
Wedge: Free pilot → prove MMR reduction → paid state contract.
Competitive Landscape
Every product on this list tells mothers what they already know. None of them tells the ASHA worker what's about to go wrong.
| Platform | Prediction | Routing | Real Govt Data | Offline | Voice |
|---|---|---|---|---|---|
| MCTS/RCH Portal | Registration only | None | Partial | No | No |
| PMSMA | Monthly checkpoint | None | No | No | No |
| ASHABot (Microsoft) | Reactive Q&A | None | No | No | Limited |
| Kilkari (BBC) | Broadcast only | None | No | SMS | IVR |
| JananiSuraksha | Predictive O(1) | Smart routing | 10,602 facilities | Yes | 12 languages |
Predictive + Routing + Tracking • O(1) • Offline-capable • Voice-first • The only one that predicts
Team
Dream. Manifest. Journey. Together as one.
Built on the clinical intelligence of vaidya-niti and the privacy architecture of mindguard. Not our first health platform.
Creator & Architect
Extending vaidya-niti (clinical decision support) and mindguard (mental health) platforms. Been there, done that.
Open to
Clinical advisors, ASHA worker partnerships, NHM collaborators
Looking for
Pilot district partnership, medical validation collaborators
Milestones & Traction
The platform is live. The engines are running. The critical next step is field validation with ASHA workers in a high-MMR district.
Three constant-time engines operational
70K risk + real facilities + 7,480 anemia profiles
Deployed on GCP Cloud Run
FastAPI + Terraform IaC, one-click deploy
34/34 tests, 6-layer security
Defense-in-depth, non-root container, audit logging
Real facilities from data.gov.in
21 states + 2 UTs, Google Maps navigation
12/12 risk factors cross-validated
NFHS-5, WHO, Cochrane, Lancet, ACOG
Voice input + Telegram alerts live
12 Indian languages, real-time family notifications
Pilot district
Q3 2026
ASHA usability study
Q4 2026
Clinical trial (RCT)
2027
Patent filing
CRI 2025
The Ask
One district. 50 ASHAs. 18 months. Rs 50 lakh. The literature-informed foundation is established. The next step is validating its effectiveness in reducing care-seeking delays.
1 District Health Officer willing to deploy with 50 ASHAs for field validation.
Obstetric specialists to confirm literature-informed thresholds against field outcomes.
18-month pilot: infrastructure + field team + cluster-randomized trial.
What we offer:
Working Platform
Deployed today
O(1) Architecture
Scales to 27M pregnancies
Potential to Reduce Delays
Contributing to preventable mortality reduction
Every mother deserves a safe delivery.
Give us one district. We aim to validate it.