Maestro Intelligence Protocol

Technical specifications for the Agentant autonomous workforce ecosystem, focusing on the MVP 1 Market Intelligence architecture.

Core Tech Stack

Frontend

  • • Next.js 14 (App Router)
  • • Tailwind CSS + Framer Motion
  • • TanStack Query

Backend

  • • Python 3.11 (FastAPI)
  • • Celery + Redis Task Queue
  • • LangChain / LlamaIndex

Data & AI

  • • PostgreSQL (Relational)
  • • Pinecone/Weaviate (Vector)
  • • Claude 3.5 / GPT-4o

Infrastructure

  • • Docker & Docker Compose
  • • Nginx Reverse Proxy
  • • Netlify / Google Cloud

Microservice Architecture

The system is composed of specialized containers orchestrated via Maestro.

/
├── apps/
│   ├── web/                     # Next.js Frontend
│   └── services/
│       ├── gateway/             # Auth, Routing, Rate Limiting
│       ├── ingestion-service/   # Data collection (Scrapers)
│       ├── intelligence-service/# AI Processing & Analysis
│       └── api-service/         # Business Logic & Data
└── infrastructure/
    └── docker-compose.yml       # Local & Prod orchestration

API Reference

MethodEndpointAction
POST/api/v1/targetsAdd company to track
GET/api/v1/targetsList tracked targets
GET/api/v1/intelligence/reportsGet AI reports
GET/api/v1/intelligence/streamReal-time alerts (SSE)

Data Strategy

Target Tracking

High-fidelity monitoring of specific keywords, competitors, and market signals stored in relational schemas for auditability.

Vector Indexing

Raw market data is vectorized and stored in Pinecone, enabling RAG (Retrieval-Augmented Generation) for accurate, grounded AI summaries.

Need custom agent integration?

Contact Engineering