Welcome to NeXifyAI
Relevant source files
The following files were used as context for generating this wiki page:
NeXifyAI is a governed AI operations and knowledge platform focused on enterprise automation. It integrates autonomous agents with a centralized "Brain" memory system to orchestrate business processes such as customer interaction and backend automation. The platform prioritizes operational excellence, scalability, and GDPR compliance, particularly for the D/A/CH B2B market.
Sources: README.md:1-5, knowledge/emergent/source/src/i18n/translations.js:14-25
System Architecture
The NeXifyAI platform utilizes a decoupled architecture consisting of a React-based frontend, a FastAPI backend, and multiple specialized database systems. It coordinates between local processing and external AI providers through an "Agent Mesh."
High-Level Component Flow
The diagram below illustrates the relationship between the user interface, the core API, and the specialized data stores.
The system routes requests through the FastAPI backend, which interacts with MongoDB for CRM data, Supabase for task orchestration, and Qdrant for semantic vector memory.
Sources: SETUP.md:92-101, services/api/routes/nexify_ai_routes.py:12-25, WORKSPACE.md:21-23
Key Service Layers
| Layer | Description | Technology |
|---|---|---|
| Frontend | Public portal, admin dashboard, and customer interface. | React 18, Vite, CSS Modules |
| Backend API | Orchestrates agent tasks, manages database state, and provides REST endpoints. | FastAPI, Python 3.11+ |
| Brain / Oracle | Centralized vector and relational knowledge base for agent context. | Qdrant (4096-dim), Supabase |
| Agent Governance | Monitors and enforces policies on autonomous AI agents. | Brain Governor, Auditor |
Sources: README.md:55-66, apps/web/README.md:9-17
Core Features
Autonomous Agent Integration
The platform uses specialized agents (e.g., AI Engineer, Quality Engineer) that execute tasks based on the "Brain-First" mandate. This mandate requires agents to query the vector memory (Qdrant) before performing actions to ensure consistency and mission alignment.
Sources: services/api/agents/skill_registry/skills/ai-engineer/SKILL.md:25-28, services/api/agents/brain_agents/documentation-expert.md:11-20
NeXifyAI Brain
The "Brain" is the company's central knowledge repository. It stores thousands of memories and knowledge chunks used for context injection during LLM calls.
- Vector Store: Qdrant handles 4096-dimensional embeddings for semantic search.
- Relational Storage: Supabase manages
oracle_tasks,brain_notes, andaudit_logs.
Sources: README.md:21-27, WORKSPACE.md:21-23
API and Integration
NeXifyAI provides a RESTful interface for both internal and external integrations. External access requires API-Key authentication via the X-API-Key header.
API Authentication Sequence
The API validates the provided key hash against the database, checks for active status, and ensures the requesting client has the appropriate scope for the endpoint.
Sources: services/api/routes/api_v1_routes.py:38-65, frontend/product-landing/src/pages/Docs.tsx:8-13
Standard Endpoints
| Endpoint | Method | Purpose |
|---|---|---|
/v1/chat/completions | POST | Generate chat completions from agent messages. |
/v1/models | GET | List available AI models and capabilities. |
/v1/contacts | GET | Retrieve a paginated list of CRM contacts. |
/api/health | GET | Check status of MongoDB, Supabase, and AI providers. |
Sources: frontend/product-landing/src/pages/Docs.tsx:23-40, services/api/routes/api_v1_routes.py:125-130
Setup and Deployment
The platform is designed to be deployed using Docker or directly on a VPS. It requires several environmental keys for operation.
Critical Infrastructure Keys
MONGO_URL: Connection string for CRM data.ALT_SUPABASE_POSTGRESQL: Database connection for the Oracle system.OPENROUTER_API_KEY: Access to LLM models (e.g.,nexify-pro).QDRANT_URL: URL for the vector memory store.
Sources: SETUP.md:14-25, services/api/routes/nexify_ai_routes.py:16-18
Security Architecture
NeXifyAI implements a Zero Trust security policy. All access requires explicit authorization, and every operation is logged in audit trails. The platform uses security tools like Gitleaks for secret scanning and Trivy for container vulnerability checks.
Sources: SECURITY.md:43-58, README.md:11-16
Summary
NeXifyAI functions as a high-performance AI ecosystem that bridges the gap between raw LLM capabilities and structured enterprise needs. By combining autonomous agent logic with the "Brain" memory system and a robust REST API, it enables scalable, governed, and compliant business automation.
Sources: README.md:1-5, services/api/routes/nexify_ai_routes.py:100-112