Skip to content

Enterprise LLM Orchestration Platform - Intelligently route requests across multiple AI providers with cost optimisation and real-time monitoring

Notifications You must be signed in to change notification settings

harshita-d/Conductor_LLM_Platform

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

59 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Multi LLM Platform

Enterprise LLM Orchestration Platform
Intelligently route requests across multiple AI providers with cost optimization and real-time monitoring.


This is an enterprise-grade LLM orchestration platform that routes requests across multiple AI providers (OpenAI, Anthropic, Google, local models), optimizing for cost, speed, and quality.

Think of it as the conductor of an AI orchestra—coordinating different AI models for the perfect performance.


🎯 Core Problems Solved

Problem Conductor Solution
🔒 Vendor Lock-in Multi-provider architecture, easy switching
💰 Unpredictable Costs Free tier optimization, real-time cost tracking
⚡ Performance Variability Intelligent routing based on request characteristics
🔍 No Visibility Monitoring & analytics dashboard
🛡️ Security Concerns Built-in governance & compliance features

🏗️ Architecture Overview

🧩 System Components

Component Purpose Technology
🖥️ Web Dashboard User Interface React + JavaScript
🚀 API Gateway Request Handling FastAPI + Python
🧠 Smart Router Provider Selection AI Logic + ML
🔌 LLM Providers AI Model Access 4 Free APIs
💾 Database Data Storage PostgreSQL
📊 Analytics Monitoring Real-time Metrics

Provider Selection Flow

Criteria Provider Description
Complex task? Gemini 🧠 Handles advanced queries

🔄 Request Flow

  1. 👤 User → API Request
    • Authentication, Rate Limiting, Validation
  2. 🧠 Smart Router → Analysis
    • Request Complexity, Speed Requirements, Provider Health
  3. 🤖 Provider Selection
    • Gemini (Quality)
  4. 📊 Response Processing
    • Analytics Logging, Cost Tracking, Performance Metrics
  5. ✅ Return to User

💸 Cost Optimization Overview

Provider Usage Bar Allocation
Gemini █████████ 100% Free (60 req/min)
Total Cost $0.00/month

✨ Key Features

🎯 Smart Routing Engine

  • Automatic provider selection based on request type
  • Performance optimization for speed, quality, or cost
  • Fallback mechanisms for provider failover
  • Custom routing rules

💰 Cost Optimization

  • Maximizes free API quotas
  • Real-time cost tracking
  • Budget alerts
  • ROI analytics

📊 Enterprise Monitoring

  • Real-time dashboards
  • Usage analytics
  • Performance benchmarking
  • Compliance reporting

🔌 Multi-Provider Support

  • Google Gemini (quality, free tier)
  • Easy extension for new providers

🚀 Quick Start

Prerequisites

  • Python 3.11+
  • Docker & Docker Compose

Requirements

  • fastapi: FastAPI is the web framework that creates your REST API endpoints.
  • uvicorn[standard]: Uvicorn is the ASGI server that runs your FastAPI application.
  • pydantic:
    • Validates that users send correct data format
    • Automatically converts JSON ↔ Python objects
    • Generates interactive API docs,
    • Catches data errors before they cause problems
    • FastAPI can work without Pydantic, but you lose most of its benefits.
    • FastAPI is designed around Pydantic and works best with it.
  • python-dotenv:
    • it's the easiest and most convenient way to manage environment variables during development.
  • google-generativeai:
    • Google Generative AI is Google's official SDK for accessing Gemini AI models.
    • Connects to Google's Gemini AI models
    • Handles authentication with Google's servers
    • Provides easy-to-use Python interface
    • Manages API calls and responses
  • httpx:
    • HTTPX is an async HTTP client for making API calls to external services.
    • good with fastAPI

folder structure

conductor-llm-platform/
├── app/
│   ├── __init__.py
│   ├── main.py
│   ├── models.py
│   └── providers/
│       ├── __init__.py
│       ├── base.py
│       └── gemini_provider.py
├── requirements.txt
├── .env
├── Dockerfile
├── docker-compose.yml
└── .dockerignore

End Points

Route Method Auth Required Purpose Status
/ GET ❌ No Welcome & info ✅ Always works
/health GET ❌ No Health check ✅ Always works
/docs GET ❌ No API documentation ✅ Always works
/redoc GET ❌ No Alt documentation ✅ Always works
/openapi.json GET ❌ No OpenAPI schema ✅ Always works
/chat POST ✅ Yes AI chat completion ⚠️ Requires providers
/status GET ✅ Yes System metrics ✅ Always works
/providers GET ✅ Yes Provider info ✅ Always works

About

Enterprise LLM Orchestration Platform - Intelligently route requests across multiple AI providers with cost optimisation and real-time monitoring

Topics

Resources

Stars

Watchers

Forks

Packages

No packages published