Transforming Insights
into Impact

A comprehensive, integrated AI platform that unifies chatbot, risk assessment, machine learning, and analytics into a single system for poverty alleviation efforts.

WorkMate brings household surveys, GATES visit logs, SBCC feedback, water and health facility data and RTV documents into one platform, where predictive models, a message generator, an assistant and dashboards turn them into six modules. The questions each module answers:

  • Who is likely to miss their income target? Risk assessment answers from household surveys, through the predictive models.
  • Where is the risk concentrated? Map analysis shows the risk predictions geographically, with water and health facilities as context.
  • How are the GATES experiments going? GATES dashboards track each experiment arm from the field visit logs. No model involved.
  • Which coaching messages actually help? SBCC analysis tests AI-written messages against field agents’ feedback.
  • What drives each prediction? ML interpretability opens the same models to show which survey features matter most.
  • How has income changed since baseline? Chat assistant writes and runs a query on survey data, with RTV documents for context.

The earlier a household is found, the more time there is to help

Raising The Village works alongside last-mile communities for 24 months, village-wide, to raise household incomes. WorkMate helps the team see early which households and clusters are on track to reach the program’s income target, and which need more support.

How Raising The Village measures impact

Six modules, one platform

Each one answers a question from the diagram above. The examples use sample data.

  1. Who is likely to miss their income target?

    Risk assessment predicts which households will hit their income target, then checks every cluster against the next survey.

    Open Risk assessment (staff sign-in required)
    Sample data
  2. Where is the risk concentrated?

    Map analysis shows the same predictions geographically, so you can see where at-risk households cluster.

    Open Map analysis (staff sign-in required)
    Sample data
  3. How are the GATES experiments going?

    GATES dashboards track the GATES experiments, comparing coaching coverage and visits across each arm.

    Open GATES dashboards (staff sign-in required)
    Sample data
  4. Which coaching messages actually help?

    SBCC analysis rates AI-written coaching messages in the field, and A/B tests how they are written.

    Open SBCC analysis (staff sign-in required)
    Sample data
  5. What drives each prediction?

    ML interpretability shows which survey features drive the model, and by how much.

    Open ML interpretability (staff sign-in required)
    Sample data
  6. How has income changed since baseline?

    Chat assistant answers from RTV’s documents and from survey data, and shows its working.

    Open Chat assistant (staff sign-in required)

    What makes a handwashing station functional?

    WorkMate

    A tippy tap or a jerrycan with a tap, placed right next to the latrine, kept full of clean water that someone checks every day, with soap or ash beside it.

    Sources · 2 RTV documents

    Sample data

Data and access

What this page shows, and who sees the rest.

Who can see household data?
Only signed-in RTV staff, and each person sees only the modules their role allows. This page shows no household records.
Are the numbers on this page real?
Only the household, cluster and district counts at the top, which come from WorkMate’s prediction data. Every other figure is sample data.
How are the AI outputs checked?
Predictions sit beside each cluster’s predicted-vs-actual results. Coaching messages are rated by field agents and A/B tested. The assistant shows its sources and the SQL it ran.

In development

Planned additions, not yet available.

Image-based evaluation
Automated classification of field photos to assess program outcomes.
Insight-triggered messaging
Targeted messages sent when the data shows a household needs support.
USSD quick help
Program information over USSD, reachable without a data connection.
Mobile companion
The assistant on a phone, with offline use for remote areas.
Partner portal
Secure access for donors and partners to follow progress and impact.

Built for the people deciding which households get support next

RTV staff

Sign in with your work account to open the modules your role can see.

Sign in to WorkMate

Partners and donors

To learn how Raising The Village uses WorkMate in its programs, talk to the team.

Contact Raising The Village