← Back to Engineering
AI / ML 2025

Mr. Muungano

A RAG-powered educator that answers Union of Tanganyika and Zanzibar questions from official sources – not model guesswork.

Role

AI Engineer

Focus

RAG product build

Stack

Python, RAG, bilingual UI

Mr. Muungano AI assistant interface

Story behind the project

Constitutions, Articles of Union, and government publications are dense. Students and young citizens need clear answers about why Tanzania has two governments, what Union Matters means, and how the Union was formed – without reading hundreds of PDF pages.

Problem statement

Public information about the Union mixes facts with opinion online, while official documents are long and hard to search. Learners also need different tones (youth, simple, student, official) and both Swahili and English.

What was built

Mr. Muungano is a retrieval-augmented generation assistant grounded in an allowlist of official sources. Users ask in natural language and get grounded answers with tone controls and a bilingual interface.

  • RAG pipeline over constitution, union documents, and curated youth Q&A
  • Source allowlist crawler to reduce speculation and rumor
  • Tone modes: youth, simple, student, and official
  • Full bilingual UI and responses (English / Swahili)

Tech used

Core tools and decisions that shaped the implementation.

Python RAG LLM APIs Official-source crawler Bilingual UI

Source & next steps

Full code, setup notes, and ongoing work live on GitHub. Open the repository to review architecture, run locally, or fork the approach.

Need a similar system?

Tell me about the problem – we can map stack, scope, and delivery.

Contact