Multi-channel intake
Clients complete intake via WhatsApp or USSD. No smartphone or internet required.
Contraceptive guidance. Delivered where women actually are.
ChaguoAI is a WHO MEC-aligned clinical decision support system serving community health workers and clients in Kenya via WhatsApp and USSD. Built by Kenyan researchers. Validated on local data. Deployed in the field.
- HASH Innovation Challenge 2026
- Data Science Africa 2026
- Maasai Mara University
Team photo placeholder — Team HEALATHTECH at Data Science Africa 2026, Makerere University, Kampala
The problem we solve
- 70%
- of global maternal deaths occur in sub-Saharan Africa (WHO, 2023)
- 218M
- women in LMICs with unmet need for modern contraception (Bearak et al., 2020)
- 35%+
- 12-month contraceptive discontinuation rate in Kenya (KDHS, 2022)
Health literacy gaps and circulating misinformation shape method choice long before a clinical conversation happens. Women arrive with fears about fertility, bleeding, and side effects that no leaflet has addressed, and often abandon a method rather than ask.
Community health workers carry the counselling load with no decision support at the point of contact. Eligibility screening against WHO criteria is done from memory, under time pressure, with paper registers and no way to check an edge case in the moment.
Digital family planning tools assume a smartphone and a data bundle. The women with the highest unmet need are reachable on basic handsets, in local languages, on channels that work offline. Most platforms simply do not reach them.
Why now
The Gates Foundation's 2026 Grand Challenges RFP states that the field lacks evidence on which AI engagement approaches improve contraceptive outcomes. ChaguoAI is not a concept. It is a deployed prototype with 78,000 training records, a validated ML model, a reproducible codebase, and a CHW pilot ready to instrument. We exist at exactly the moment this RFP is asking for.
How the system works
WHO MEC safety engine
A deterministic rule engine maps each client profile to WHO eligibility categories before any recommendation is made.
RAG-powered counselling
Retrieval from Kenya MOH guidelines and WHO publications grounds every recommendation in cited, vetted guidance.
CHW confirmation
Community health workers review, counsel, and confirm every client choice. ChaguoAI supports, it does not prescribe.
- Intake (WhatsApp / USSD)
- MEC Engine
- RAG + LLM
- Recommendation Packet
- CHW Review
- Client Confirmed Choice
- Follow-up
Our technology
| Layer | Technology |
|---|---|
| Safety engine | WHO MEC rules (deterministic, not learned) |
| Knowledge retrieval | ChromaDB + OpenAI embeddings |
| Language generation | Google Gemini 2.5 Flash |
| Discontinuation risk | LightGBM classifier (AUC-ROC 0.903) |
| Channels | WhatsApp (Twilio), USSD (Africa's Talking) |
| Backend | Flask, Redis Queue, Firestore |
| Deployment | Render (cloud) |
The ML model was trained on 78,000+ anonymised family planning programme records from Siaya and Busia counties in Western Kenya, sourced from the DASSA platform under the HASH Innovation Challenge. It predicts 12-month discontinuation risk with an AUC-ROC of 0.903 and an Expected Calibration Error of 0.005. The model annotates recommendation packets in shadow mode only. WHO MEC eligibility rules always take precedence. Clinical safety is never ML-determined.
Source code is MIT-licensed and publicly available. Full reproducibility package provided.
View RepositoryThe system in action
Four interfaces. One coordinated care workflow.
Screenshot: WhatsApp conversation flow — upload image here
Client intake via WhatsApp
Clients initiate a structured intake conversation through WhatsApp. The system collects age, parity, medical history flags, and method preferences across multiple turns. Responses are generated in plain language grounded in Kenya MOH guidelines. No app download required.
- Multi-turn conversation
- Swahili and English
- No smartphone needed
- Twilio-powered
Screenshots will be updated as the field pilot progresses. To request a live demo of the system, contact the team using the form below.
Request a demoRecognition and validation
June 2026 · HASH Innovation Challenge
Best solution, Track II — Contraception Decision-Making
ChaguoAI was selected as the leading prototype in a competitive field of regional teams addressing contraceptive decision support across sub-Saharan Africa.
July 2026 · Data Science Africa 2026, Makerere University, Kampala
Poster and pitch presentation
Team HEALATHTECH presented ChaguoAI to an international audience of data scientists, public health researchers, and funders at DSA 2026.
August 2026 · Gates Foundation Grand Challenges RFP
Applicant — AI-Enabled Family Planning Consumer Engagement
ChaguoAI is applying for a Gates Foundation Grand Challenges grant to scale evidence-based AI contraceptive counselling across Kenya.
"What the field lacks is evidence. We do not yet know which AI-enabled engagement approaches work best."
ChaguoAI is designed to answer exactly that question, with reproducible methodology and locally grounded data.
Built by Kenyans, for the region
Godfrey [Last name]
Principal Investigator, Technical Lead
CS finalist, Maasai Mara University. Founder, ZelvoFlow. Builder of DriveCare and ChaguoAI.
Felistus Mukhwana
Software Engineering
Backend architecture, Flask, Firestore, Redis, CI pipeline.
Moses Maemba
Clinical and Public Health
WHO MEC alignment, Kenya MOH guideline mapping, clinical guardrails.
Hannah Njeri
UX/UI and Nursing
Provider portal, WhatsApp and USSD conversation flows, low-literacy accessibility.
We are actively seeking institutional collaborators, clinical advisors, and implementation partners in Kenya and East Africa.
0+
Training records from Western Kenya
0.000
Model AUC-ROC on holdout test set
0
Offline unit tests in CI
0
Channels: WhatsApp, USSD, CHW portal, admin dashboard
Institutions and platforms
Maasai Mara University
HASH Innovation Challenge
Data Science Africa
DASSA Platform
[Placeholder: Your NGO Partner]
[Placeholder: MOH Kenya]
[Placeholder: Add partner]
We are seeking MOH Kenya endorsement and pilot implementation partnerships with county health systems in Western Kenya.
Press and recognition
Data Science Africa · July 2026
ChaguoAI presented at DSA 2026, Makerere University
Team HEALATHTECH showcased a WHO MEC-aligned decision support system for family planning at the annual DSA gathering.
Add outlet name · Add date
Add press mention here
Add press mention here.
Add outlet name · Add date
Add press mention here
Add press mention here.
Get involved
Fund the next phase
We are applying for the Gates Foundation Grand Challenges AI-Enabled Family Planning grant. If you represent a foundation, bilateral, or impact investor interested in evidence-based AI for reproductive health, we want to hear from you.
Contact the PIPartner on implementation
We are seeking county health system partners, NGOs, and digital health platforms with existing FP programme reach in Kenya and East Africa. The CHW portal is ready for pilot deployment.
Discuss a PartnershipUse our evidence base
Our interaction data, conversation quality rubrics, and bias audit methodology will be made available to the field. If you are a researcher or evaluator working on AI-enabled health engagement, reach out.
Request AccessSupport the work
ChaguoAI is an open-source, community-built system. We are a team of four Kenyan researchers and clinicians building technology that serves women who have no other access to reliable contraceptive guidance.
Your contribution funds server costs, clinical data validation, field testing in Western Kenya, and the time of the CHW workers who pilot the system. Every dollar goes directly to making the prototype field-ready.
ChaguoAI is a research prototype operated by Team HEALATHTECH. For institutional giving or grant co-funding inquiries, use the contact form above.