Aman Jain

AI Analyst | Agentic AI & LLM Applications | LangGraph, RAG, MCP, FastAPI

I am an AI Analyst at Impact Analytics and a Computer Science graduate from NIT Raipur, working at the intersection of data, forecasting, analytics engineering, and applied AI systems.

At work, I focus on data quality, implementation readiness, forecasting workflows, SQL debugging, pipeline failure triage, cross-platform validation, and client-facing analytics for large retail businesses. I enjoy taking ambiguous requirements, breaking them down into reliable technical workflows, and driving them to production-ready delivery.

Alongside my professional work, I am building deeply in Agentic AI Engineering, especially:

• LLM applications with tool calling and memory
• LangGraph-based agent workflows
• Retrieval-Augmented Generation (RAG)
• Model Context Protocol (MCP) servers
• FastAPI-based AI backends
• Vector search and source-grounded retrieval
• Docker/AWS deployment workflows
• Evaluation, observability, and guardrails for AI systems

I am especially interested in roles where I can combine software engineering, data systems, AI agents, strong debugging, and forward-deployed problem solving to build systems that create real business value.

• Building production-style agentic AI applications with LangGraph, RAG, memory, and tools
• Creating MCP servers that expose reliable external capabilities to AI applications
• Improving practical depth in LLM evaluation, tracing, guardrails, and deployment
• Strengthening backend and data engineering foundations for real-world AI systems

AI / LLM
LangGraph LangChain MCP FastMCP RAG Tool Calling Prompt Engineering Embeddings Vector Search Gemini Groq OpenAI API Tavily Serper

Backend / Data
Python FastAPI SQL PostgreSQL Snowflake BigQuery Airflow GCS SQLAlchemy SQLite MongoDB MongoDB Atlas ChromaDB

Delivery / Debugging
Data Validation Data Quality Pipeline Debugging SQL Optimization UAT Client Delivery Stakeholder Communication Cross-functional Collaboration

Cloud / DevOps
Docker GitHub Actions Amazon ECR EC2 REST APIs Postman Git

ML / CV
scikit-learn YOLOv8 OpenCV Pandas NumPy

Repository: github.com/amanjain200/agentic_chatbot

A production-style agentic chat application with a FastAPI backend, browser-based chat workspace, LangGraph orchestration, Gemini models, tool calling, document RAG, persistent memory, and Docker/AWS deployment workflow.

Highlights:

• LangGraph agent loop with tools for web search, calculator execution, uploaded-document retrieval, memory save, and memory recall
• Thread-scoped RAG over PDF, DOCX, TXT, Markdown, Python, and CSV uploads
• Chroma vector search with Gemini embeddings
• SQLite persistence for conversations, messages, memories, and LangGraph checkpoints
• Dockerized app with GitHub Actions CI/CD to Amazon ECR and EC2
• Roadmap: auth/RBAC, streaming responses, eval harness, observability traces, guardrails, MCP/A2A connectors, and live demo deployment

Repository: github.com/amanjain200/official-docs-mcp

A Python MCP server that exposes a documentation retrieval tool for AI agents, helping agents answer developer questions using source-grounded context from official documentation.

Highlights:

• Built with FastMCP over stdio
• Exposes a get_docs tool for trusted documentation lookup
• Supports official docs for LangChain, LlamaIndex, OpenAI, and uv
• Uses Serper search, async httpx, and trafilatura page extraction
• Includes an MCP client that discovers tools, invokes the server, and passes source-labeled context to a Groq-hosted LLM
• Includes a context-sharing MCP server example for multi-agent handoff patterns

AI Analyst, Impact Analytics
Working on data quality, forecasting workflows, implementation readiness, SQL debugging, validation automation, and cross-functional delivery for large retail clients.

Selected highlights:

• Owned data validation for a client go-live that finished with zero data-related tickets after launch
• Built cross-platform cloud data integration and validation workflows for downstream client requirements
• Debugged complex SQL defects and pipeline failures across warehouse tables, transformations, scheduled workflows, and client-facing outputs
• Designed an incremental forecast reporting framework using prior forecasts, refreshed forecasts, and actuals in a running-window structure
• Recognized with company Quarter Awards in Q2 2025 and Q1 2026, plus the Gen AI Pacesetter Award in Q2 2025

• Agentic AI - Skill Up - GeeksforGeeks
• Postman API Fundamentals Student Expert - Canvas Credentials (Badgr)

• Email: [email protected]
• LinkedIn: linkedin.com/in/the-aman-jain

_{Building practical AI systems with agents, data, tools, memory, retrieval, and production-minded engineering.}