langraph-python

Local-first RAG pipeline using TurboVec for vector search and LangGraph for orchestration, with Gemini API for response generation

This is a LangGraph AI Agent that features a sophisticated two-tier routing system that uses OpenAI GPT models for smart tool selection with rule-based pattern matching as a reliable fallback, making it production-ready with extensive testing coverage and interactive capabilities.

Tree Evaluator is a minimal FastAPI application that computes above- and below-ground biomass, carbon, and CO₂ from dendrometric parameters (e.g. diameter, height), using an explicit OOP calculation service.

This project is a Mini Agentic RAG (Retrieval-Augmented Generation) System designed to answer domain-specific questions with high accuracy and minimal hallucinations. It leverages Azure OpenAI for LLM capabilities and FAISS for efficient vector retrieval.

Multi-agent Lang-graph framework for autonomous software developement

Agente de IA especializado en estudios de mercado y análisis financiero cuantitativo para Colombia y Latinoamérica.

AI Agent to Handle your Time and Your tasks (Intelligence TODO ) over telegram bot.

Real-time API monitoring powered by GenAI. Detect anomalies, track LLM costs, and search logs semantically with Pinecone and Groq.

Pedagogical LangGraph project: human-in-the-loop, tool agents, and reflection patterns

Project to demonstrate MCP and A2A agentic agent technology

An AI agent to automate daily job applications across multiple platforms, reducing manual effort by Implementing LangGraph & Groq API with scheduling, recruiter notification system, and automated resume parsing and multi-agent workflow,

LLM evaluation lab: offline evals, LLM-as-a-judge, and prompt versioning — built on a real-world ReAct agent with RAG and LangGraph

Agentic AI Assistant for Financial Market Research

Documentation Agent for GeniOS Ecosystem. This agent helps create Standard Operation Procedure (SOP) Document through Email

🩺 Agentic RAG chatbot for medical Q&A on Diabetes, Hypertension & Asthma. Powered by a LangGraph multi-step reasoning agent, hybrid retrieval (dense + BM25 + Cohere rerank), Groq LLaMA 3.3 70B, Qdrant, and Supabase. FastAPI backend on Cloud Run.

A multi-agent system that automatically reads ML research papers, analyses their code, samples datasets, runs experiments, and writes a structured report to Notion — triggered by n8n and managed through a web UI.

This project presents a production-grade multi-agent research system that automates the end-to-end workflow of technical research: searching the web via DuckDuckGo, extracting structured information, cross-verifying sources, and synthesizing a structured JSON report — all powered by Groq's LPU inference on Llama 3.3 70B.

LangGraph agent that turns natural language into Beancount transactions with validation and human-in-the-loop confirmation via Chainlit.

Transforming educational content into classroom-ready teaching intelligence.