evidoria

← Back to browse

Good practice Imported

GoR AI-Powered Policy Navigator — Rwanda's RAG Assistant for Government Policy Access

Rwanda · Kigali · See the Rwanda profile · See the Kigali profile

Evidence: Descriptive / self-reported Top 67% 53/100 · Ask Evidence Copilot about this practice

Built by RISA with NISR, AIMS Rwanda and Cenfri, Rwanda's Policy Navigator uses retrieval-augmented generation to answer questions on government policy from verified official documents, and won recognition among 59 of ~400 entries at ITU's 2026 AI for Good awards.

50
Professionals trained in the DSCBI first cohort
24
Government institutions represented in the DSCBI cohort
1 of 59 (from ~400 submissions)
Recognition among global submissions at ITU AI for Good 2026
GoR AI-Powered Policy Navigator — Rwanda's RAG Assistant for Government Policy Access

Details

Maturity
Pilot
Promoter
Rwanda Information Society Authority (RISA)
Period
2025-2026
Keywords
digital government, policy analysis, public administration

Context

Civil servants and citizens in Rwanda often struggle to locate the specific, current version of a policy, regulation or strategy among scattered government documents, slowing both compliance checks and public information requests. The Rwanda Information Society Authority (RISA), working with the National Institute of Statistics of Rwanda (NISR), the African Institute for Mathematical Sciences - Rwanda (AIMS Rwanda), Mastercard and Cenfri, built the 'GoR AI-Powered Policy Navigator'.

Objectives

The tool uses retrieval-augmented generation (RAG) to answer multilingual queries about policies, regulations and strategies, drawing responses exclusively from verified government sources, and includes a secure workspace and an automated policy-alignment checker against national frameworks.

Activities

The Policy Navigator was built during the first cohort of the Data Science Capacity Building Initiative (DSCBI), a programme that trained 50 professionals from 24 government institutions, and was one of the top three projects from that cohort.

Results

The project was one of 59 projects, from nearly 400 global submissions, recognised at the ITU AI for Good Global Summit's Innovate for Impact 2026 awards, in the Open-Source AI for Global Impact category.

Conclusions

This is external, competitive validation of the concept, but neither RISA's announcement nor the award materials publish usage volumes, response-accuracy figures or a comparison of search time before and after; the efficiency claim that the tool 'significantly reduces time required to search through policy documents' remains self-reported.

Implementation
Indicative cost
Low (< €50k)
Time to results
Short (< 1 year)
Staffing & skills
Rwanda Information Society Authority (RISA), lead, National Institute of Statistics of Rwanda (NISR), African Institute for Mathematical Sciences - Rwanda (AIMS Rwanda), Mastercard, Cenfri

Conditions for success

  • Data Science Capacity Building Initiative (DSCBI) training 50 professionals from 24 institutions
  • RAG architecture restricted to verified government sources
  • Secure workspace and automated policy-alignment checker

Common failure modes

  • No usage-volume, accuracy or before/after search-time figures published; the efficiency claim is self-reported

Commonly funded by

National / regional programmes Philanthropic / foundation funding

Indicative funding routes for practices of this type — always check each programme's current calls and eligibility rules.

Do you run this practice? Claim it — verified implementers get a public contact pathway and can propose corrections.

Data sources

Where this practice's information was retrieved from, and when.

Attachments

Similar practices you may find useful