Bops
AI 봇 팀이 업무 운영을 담당, 메일·파이프라인·청구·보고서 관리, 각각 컴퓨터·이메일·전화번호 보유, 모든 정보를 기억
Facil: 적도기니아 정부 디지털 서비스 프레임워크, CI/CD 현황·아키텍처 개요·프로젝트 마일스톤을 보여주는 플랫폼 대시보드
Facil: Government Digital Services Framework for Equatorial Guinea. Platform monitoring dashboard with CI/CD status, architecture overview, and project milestones.
비즈니스PLpgSQL · TypeScriptGitHub @KouemouSah온라인운영자 미확인
써 보셨어요? 한 번만 눌러 주세요. 로그인은 필요 없어요.
Facil — Scalable AI-Powered Government Digital Services Framework. Document intelligence, workflow automation, compliance monitoring. Currently deployed for Equatorial Guinea.
README 전체를 불러오는 중입니다. 아래는 저장된 발췌입니다.
Powering digital transformation for government procedures, workflows, and fiscal services — deployable to any country.
Currently deployed for the Republic of Equatorial Guinea.
Facil is a full-stack, AI-powered framework for digitizing government administrative procedures. It provides a complete platform — backend API, web dashboard, citizen mobile app, and field inspection app — that any government can configure to manage its own fiscal services, workflows, document processing, and citizen interactions.README 전문 보기 ↗Powering digital transformation for government procedures, workflows, and fiscal services — deployable to any country.
Currently deployed for the Republic of Equatorial Guinea.
Facil is a full-stack, AI-powered framework for digitizing government administrative procedures. It provides a complete platform — backend API, web dashboard, citizen mobile app, and field inspection app — that any government can configure to manage its own fiscal services, workflows, document processing, and citizen interactions.
The framework is country-agnostic by design: workflows, fiscal services, entities, roles, translations, and business rules are all data-driven and configurable through the database and admin UI — not hardcoded. Deploying Facil to a new country means configuring its services, entities, and workflows, not rewriting code.
Current deployment: Republic of Equatorial Guinea — 873 fiscal services, 21 ministries, 20 government entities, 36 business workflows, trilingual (Spanish, French, English).
End-to-end digitization of administrative procedures. 36 predefined workflows (passport, residency, driving license, contracts, inspections...) with configurable steps, dynamic form validation, conditional routing, and multi-entity approval chains. Citizens submit requests through guided wizards; government agents process them via role-based dashboards with SLA-driven priority queues, automatic assignment, and real-time workload balancing across 20 government entities.
A RAG chatbot powered by Gemini 2.5 Flash with role-based specialized agents — each agent has domain-specific tools and access controls respecting strict data confidentiality:
| Agent | Scope | Tools | Confidentiality |
|---|---|---|---|
| Citizen Agent | Public services, fees, procedures | 19 public tools | No access to internal data |
| Authenticated Agent | Personal requests, status tracking | 8 auth tools | Own data only |
| Treasury Agent | Payment validation, revenue analysis | Deep reasoning | Entity-scoped financial data |
| Supervisor Agent | Team performance, anomaly detection | Deep reasoning | Entity-scoped agent metrics |
| Admin Agent | System configuration, audit analysis | Deep reasoning | Full system access with audit trail |
Each agent operates with Chain-of-Thought reasoning (6-step prompt), self-reflection (auto-regeneration if quality score < 5/10), and 15 response formats.
Convert physical documents (contracts, invoices, identity papers, government forms, tax declarations) into structured digital data. The OCR pipeline extracts, validates, and structures data from 39+ document schemas using template-based field mapping with bounding box annotation, AI-powered classification, and cross-document validation. Extract structured text from PDFs while preserving layouts.
Every agent action, payment validation, document review, and permission change is logged in the audit trail (2,800+ entries). The SchemaValidationEngine enforces document integrity with 70+ JSON rules. RiskAnalyzer runs a 12-step assessment pipeline on every document. All API requests pass through rate limiting, JWT validation, RBAC check (335 permissions), and parameterized SQL — ensuring OWASP compliance at every layer.
AI-driven analysis: detect recurring patterns across scanned documents, validate data by cross-referencing multiple sources (OCR extraction vs form input vs database records), and enable enterprise-grade search across the fiscal services catalog (873 services, 21 ministries). Automatic zone resolution with Levenshtein fuzzy matching and commerce type classification for commercial license workflows.
| Metric | Value |
|---|---|
| Fiscal Services Catalog | 873 (869 active) |
| Government Entities | 20 |
| Business Workflows | 36 predefined |
| Database Tables | 145 |
| Database Enums | 50 |
| Roles | 47 (335 granular permissions) |
| OCR Document Schemas | 39 |
| Languages | 3 (ES, FR, EN — extensible) |
| Backend Modules | 30 |
| Frontend Modules | 41+ |
| Communication Templates | 85 (29 email, 31 SMS, 9 push, 16 in-app) |
| Audit Log Entries | 2,800+ |
| Total Commits | 3,800+ |
여기까지가 README 앞부분입니다. 나머지는 원문에서 읽을 수 있습니다. README 전문 보기 ↗