I design and ship production multi-agent systems — from architecture to evaluation. My day-to-day is deep in agent orchestration with LangChain/LangGraph, building and integrating MCP servers, crafting precise system prompts, and running agent evaluation pipelines to ensure reliability before anything hits prod.
I care about agents that actually work — not just demos. That means robust tool design, deterministic routing, failure handling, and measurable quality.
| Area | What I work on |
|---|---|
| Agent Architecture | Multi-agent pipelines, autonomous agents, tool use, LangGraph state machines |
| MCP Servers | Custom MCP server development, tool schema design, enterprise integrations |
| Prompting | System prompt engineering, chain-of-thought, structured output, prompt versioning |
| Agent Evaluation | Eval framework design, LLM-as-judge, regression testing, tracing & observability |
| LLM Infrastructure | RAG pipelines, vector DBs, embedding strategies, retrieval optimization |
Python LangChain LangGraph FastAPI MCP OpenAI Anthropic Claude
RAG Vector DBs Prompt Engineering Agent Evals Docker REST APIs
I contribute to projects where I can add real, lasting value — not just quick fixes. Areas I focus on:
- 🔌 MCP servers & tooling — custom integrations, tool schema design, server patterns
- 🧩 Agent frameworks — LangGraph, LangChain, evaluation tooling
- 📊 LLM observability & evals — tracing, scoring, reliability testing
- 🗂️ RAG & retrieval infrastructure
If you're a maintainer in these spaces and want a contributor who understands production constraints — let's connect.
| Project | Description |
|---|---|
| AutoInsight-AI | LLM-powered automated data analysis and insight generation |
| Invoice Extraction Pro | Intelligent document parsing with vision models and structured extraction |
| Stack Chatbot AI | RAG-based Q&A system over technical documentation |
| Intelligent Inventory Management | End-to-end AI-driven inventory forecasting and management |
Remote-first Senior AI Engineer roles in multi-agent systems, LLM infrastructure, or AI platform engineering.
📬 akshaykumarbedre.bm@gmail.com