Projects/Peregrine Compass - MongoDB BI LLM
backend
AI/ML
NDA Protected

Peregrine Compass - MongoDB BI LLM

Natural Language to MongoDB Query Generator

Overview

Production-ready system for converting natural language queries to MongoDB aggregation pipelines using fine-tuned LLMs. Powered by Ollama for local-first AI without API dependencies.

Key Features

  • 100% local AI processing (no cloud dependencies)
  • Fine-tuned LLMs for database queries
  • Natural language understanding
  • MongoDB aggregation pipeline generation
  • RAG (Retrieval-Augmented Generation) support
  • FastAPI REST endpoints
  • Docker containerization
  • Query execution and result streaming

Outcomes & Impact

  • Reduced query writing time by 80%
  • No API costs - 100% self-hosted
  • Production-ready error handling
  • Scalable to enterprise workloads

Tech Stack

Python

backend

FastAPI

backend

Ollama

other

MongoDB

database

Mistral LLM

other

RAG

other

Docker

devops

Tags

Python
Ollama
MongoDB
FastAPI
LLM
RAG
Mistral

Project Info

Categories

backend, AI/ML

Status

Confidential