Logo Lanfrica

gabrielmahia/taarifa-ai

Domaine:

natural language processing

Type de record:

software
Créateur:
gab
Hôte:
TaarifaAI — AI-powered civic journalism briefing for Kenya. Aggregates news, identifies narratives, flags institutional accountability gaps. # 📰 TaarifaAI — Kenya Civic Briefing AI > *Taarifa* (Kiswahili) — report, news, briefing. AI-powered civic briefing tool for Kenya. Feed it a county, a time period, or a topic — get a structured investigative brief: budget anomalies, procurement gaps, parliamentary performance, and constitutional rights violations — in plain language, ready to publish. ## Who this is for - **Investigative journalists** covering devolution, procurement, and public finance - **Civil society organisations** doing budget accountability work - **County assemblies** wanting to scrutinise executive spending - **Academics and researchers** studying Kenya's devolution - **Citizens** who want to understand how their county's money is spent ## What it produces Given a county and FY, TaarifaAI generates a structured brief: ``` COUNTY ACCOUNTABILITY BRIEF — TURKANA, FY 2022/23 Generated: April 22, 2026 | Source: Controller of Budget + OCDS BUDGET SUMMARY Development allocation: KES 4.2B Absorbed: KES 1.72B (41%) Unspent: KES 2.48B — flagged for follow-up PROCUREMENT CROSS-CHECK Contracts in OCDS: 8 tenders, 5 awarded Total contract value: KES 890M Unexplained gap: KES 830M — no procurement records KEY QUESTIONS FOR INVESTIGATION 1. Where did KES 830M in absorbed spend go without tender records? 2. 3 road contracts awarded to single vendor — possible single-sourcing? 3. 0 health facility tenders despite KES 400M health allocation CONSTITUTIONAL ANGLE Article 201: Public finance must promote accountability Article 35: Citizens have the right to access this information RECOMMENDED NEXT STEPS → File FOI request under Access to Information Act 2016 → Contact County Assembly Budget Committee → Cross-reference with IFMIS records ``` ## Quickstart ```bash git clone github.com cd taarifa-ai pip install -r requirements.txt streamlit run app.py ``` ## Data sources - **Controller of Budget** — county budget …