Skip to content

Releases: anderson-ntlabs/cidadao.ai-backend

Release list

v1.0.0 - Initial Public Release

v1.0.0 - Initial Public Release Latest
Latest

Choose a tag to compare

View all tags
@anderson-ntlabs anderson-ntlabs released this 12 Dec 13:16
v1.0.0
05d114f

Cidadão.AI Backend - Initial Public Release

First public release of the Cidadão.AI Backend - a multi-agent AI system for Brazilian government transparency analysis.

Features

  • 17 specialized AI agents named after Brazilian historical figures
  • FastAPI REST API with comprehensive documentation
  • Anomaly detection in government contracts
  • Integration with Brazilian transparency portals
  • JWT authentication and enterprise security

Documentation

  • Full API documentation at /docs
  • CITATION.cff for academic citations
  • MIT License

How to Cite

SILVA, A. H. (2025). Cidadão.AI Backend: Multi-Agent AI System for Brazilian Government Transparency Analysis. GitHub. https://github.com/anderson-ufrj/cidadao.ai-backend

ORCID: https://orcid.org/0009-0001-7144-4974

Assets 2
Loading

V1.0.0-beta - First Production Beta Release 🚀

V1.0.0-beta - First Production Beta Release 🚀 Pre-release
Pre-release

Choose a tag to compare

View all tags
@anderson-ntlabs anderson-ntlabs released this 18 Nov 20:02
v1.0.0-beta
940d411

🎉 V1.0.0-beta - First Production Beta Release

Official beta launch of Cidadão.AI Backend - Production-ready multi-agent AI system for Brazilian government transparency analysis.

After 3 months of intensive development (August-November 2025), 1,080 commits, and comprehensive testing, the system is ready for public beta testing.

🏆 Production Metrics

Metric Value Status
Test Coverage 76.29% ✅ Above 75% target
Tests 1,514 (97.4% pass) ✅ High quality
Production Uptime 99.9% ✅ Excellent
E2E Tests 5/5 passing ✅ 100% success
Response Time 0.6s average ✅ 70% better than target
Agents Operational 16/16 ✅ All tiers functional

✨ Key Features

🤖 Multi-Agent System (16 Agents)

  • Tier 1 - Excellent (10 agents, >75% coverage):

    • Zumbi dos Palmares (96.32%) - Anomaly Detection
    • Anita Garibaldi (94.87%) - Pattern Analysis
    • Oxóssi (94.44%) - Data Hunting
    • Lampião (93.75%) - Regional Analysis
    • Ayrton Senna (92.31%) - Semantic Routing
    • Tiradentes (91.67%) - Report Generation
    • Oscar Niemeyer (89.47%) - Data Aggregation
    • Machado de Assis (88.24%) - Textual Analysis
    • José Bonifácio (87.50%) - Legal Analysis
    • Maria Quitéria (86.96%) - Security Auditing

  • Tier 2 - Near-Complete (5 agents):

    • Abaporu (85.71%) - Master Orchestration
    • Nanã (84.62%) - Memory Management
    • Drummond (83.33%) - Communication
    • Céuci (82.76%) - ETL & Predictive Analytics
    • Obaluaiê (81.25%) - Corruption Detection

  • Tier 3 - Framework (1 agent):

    • Dandara (86.32%) - Social Equity Analysis

🌐 Government Data Integration

  • 30+ Brazilian government APIs integrated
  • Federal APIs (8): IBGE, DataSUS, INEP, PNCP, Compras.gov, Portal da Transparência, Banco Central, Minha Receita
  • State APIs (5): TCE-CE, TCE-PE, TCE-MG, SICONFI, CKAN
  • Real-time data federation with circuit breakers

⚡ Performance Optimizations

  • Agent Lazy Loading: 367x faster imports (1460ms → 3.81ms)
  • Multi-Layer Caching: ~95% hit rate (Memory → Redis → Database)
  • Connection Pooling: PostgreSQL (20 connections) + Redis (50 connections)
  • Compression: Gzip + Brotli middleware

🔒 Security & Infrastructure

  • JWT-based authentication + API keys
  • Rate limiting per user/IP
  • CORS, CSRF, XSS protection
  • PostgreSQL + Redis on Railway
  • Prometheus metrics + Grafana dashboards

🚀 Try It Now

Production Endpoints

Quick Start

```bash

Clone repository

git clone https://github.com/anderson-ufrj/cidadao.ai-backend
cd cidadao.ai-backend

Checkout v1.0.0-beta

git checkout v1.0.0-beta

Install dependencies

pip install -r requirements.txt

Configure environment

cp .env.example .env

Add MARITACA_API_KEY or ANTHROPIC_API_KEY

Run development server

python -m src.api.app

API available at http://localhost:8000

```

📊 System Status

Component Readiness Notes
Backend ✅ 100% All must-have criteria met
Frontend ✅ 90% Deployed on Vercel, chat working
Integration ✅ 85% End-to-end functional
Documentation ✅ 100% Comprehensive and complete

🔧 Technical Highlights

Performance Benchmarks

Metric Target Actual Status
API Response (p95) <2000ms ~600ms ✅ 70% better
Agent Processing <5000ms ~3200ms ✅ 36% better
Chat First Token <500ms ~380ms ✅ 24% better
Investigation (6 agents) <15000ms ~12500ms ✅ 17% better
Agent Import Time <100ms 3.81ms ✅ 96% better

Architecture

  • Orchestrator Pattern: Coordinated agent execution
  • Reflection Pattern: Self-improving agent quality
  • Entity Graph: NetworkX-based relationship tracking
  • Data Federation: Parallel API calls with fallback
  • Circuit Breakers: Resilient external API integration

📚 Documentation

New in V1.0

Production Validation Reports

Architecture & Guides

⚠️ Known Limitations

Expected (Non-Blocking)

  1. Portal da Transparência: 78% of endpoints return 403 (government API restriction)
  2. Load Testing: Not yet performed (needs real production traffic)
  3. Production Alerting: Grafana alerts configured but not deployed
  4. Advanced Observability: Distributed tracing (Jaeger) planned for V1.1

None of these block V1.0 launch - all are post-launch optimization tasks.

🎯 Roadmap

V1.1 (December 2025)

  • OAuth social login
  • WebSocket real-time updates
  • Performance optimization based on real traffic
  • Grafana production alerts
  • Load testing and capacity planning

V2.0 (Q1 2026)

  • ML models custom-trained on Brazilian data
  • Predictive analytics
  • Advanced visualizations
  • Multi-tenancy support
  • Enterprise features

🤝 Contributing

We welcome contributions! See CONTRIBUTING.md for guidelines.

Quick Contribution Steps

  1. Fork the repository
  2. Create feature branch (`git checkout -b feature/amazing-feature`)
  3. Make changes and test (`make test`)
  4. Commit (`git commit -m 'feat: add amazing feature'`)
  5. Push (`git push origin feature/amazing-feature`)
  6. Open Pull Request

🙏 Acknowledgments

Brazilian Cultural Icons who inspired our agent identities:

  • Zumbi dos Palmares, Anita Garibaldi, Tiradentes, Ayrton Senna, and 12 more incredible Brazilians

Open Source Community:

  • FastAPI, Pydantic, SQLAlchemy, and many more amazing projects

Brazilian Government:

  • For open data initiatives making transparency accessible

📞 Contact & Support

Primary Developer: Anderson Henrique da Silva (Minas Gerais, Brasil)

Links:

📈 Development Statistics

  • Development Period: August 13 - November 18, 2025 (97 days)
  • Total Commits: 1,081
  • Lines of Code: ~133,783 (src/)
  • Test Code: ~49,888 lines
  • Test Files: 98 files
  • Contributors: 1 (open for more!)

🇧🇷 Made with ❤️ in Minas Gerais, Brasil

🚀 Democratizing Government Transparency Through AI

Release Date: November 18, 2025
Version: 1.0.0-beta
Status: Production Ready - Beta Testing

For full details, see CHANGELOG.md

Loading