Hi, I'm

Tanvir Ahmod Rafi

AI Engineer

I build production-grade AI systems, automated chatbots, and scalable backend services with RAG, MCP tool orchestration, vector databases, n8n workflow automation, and API-first architecture.

Explore my projects

About

Open to work

I'm Tanvir Ahmod, an AI Backend Engineer focused on intelligent backend systems, chatbot infrastructure, and scalable AI application development.

I specialize in Retrieval-Augmented Generation (RAG), MCP-based orchestration, vector search, FastAPI services, and n8n workflow automation to build practical AI products that solve real-world problems. I'm open to full-time and contract opportunities (onsite / remote).

397 contributions in 2025 on GitHub.

LessMore

GitHubtanvirahmodrafi

LinkedIntanvirahmodrafi

Mail Metanvirahmod.cn@gmail.com

Tools & Technologies

My stack centers on AI backend engineering, retrieval systems, APIs, and full-stack product delivery.

Python

TypeScript

LangChain

LangGraph

MCP

n8n

PyTorch

scikit-learn

NumPy

FastAPI

Node.js

NestJS

Next.js

React

Prisma

Zod

PostgreSQL

MongoDB

Redis

Supabase

Qdrant

Pinecone

Git

Docker

Experience

AI Analyst
Anwar Group of Industries•Full-time
Jul 2026 - Present
Jul 2026 - Present
Added a module to Dristi — the app that manages the A1 Polymer business (serving around 700 users) — that automatically sends SMS updates summarizing each responsible person's work using n8n.

Automated WhatsApp workflows using n8n as the AI backend.

Developed the UI and backend with JavaScript and PHP, backed by MySQL.

Designed and deployed the complete end-to-end automation pipeline for the module.
- n8n
- JavaScript
- PHP
- MySQL
- WhatsApp API
AI Backend Engineer
Saratech Ltd•Full-time
Dec 2024 - Jun 2026
Dec 2024 - Jun 2026
Developed production-ready AI backend systems and chatbot infrastructure using FastAPI, LangChain, and LangGraph.

Built and maintained Ulka Chat, implementing RAG-based retrieval pipelines, vector search workflows, and context-aware response generation.

Integrated MCP, Qdrant, and LLM-based services with scalable API architecture and modular tool orchestration.
- Python
- FastAPI
- RAG
- MCP
- LangChain
- LangGraph
- Qdrant
Founding Member & Vice Chair
IEEE CIS — SEU Student Branch Chapter•Volunteer
2026 - Present
2026 - Present
Research and Development Wing Member
IEEE CS — SEU Student Branch Chapter•Volunteer
2025 - Present
2025 - Present

Education

Southeast University
B.Sc. in Computer Science & Engineering · CGPA: 3.90
Oct 2022 - Oct 2026
Rajshahi Education Board Govt College
Higher Secondary Certificate (HSC)
2019 - 2021
Harimohan Government High School
Secondary School Certificate (SSC)
2014 - 2019

Projects

VCP Winner

RFPilot AI

VCP-winning AI proposal command center — shipped with just 5 professional prompts on Lovable, in under 60 minutes. Paste raw RFP text and it extracts every must/shall/should clause in under 10 seconds, drafts proposal-grade sections in ~8s each, scores readiness live on a 0–100 scale, and tracks submissions across a Discovery → Drafting → Review → Won Kanban with win-rate analytics.

React
TypeScript
Vite
Tailwind CSS
Supabase
LLMs

Live Demo View Github

AGI Logistics

End-to-end inbound transport procurement for Anwar Group, replacing the nightly WhatsApp scramble with a timed reverse auction. Vendors bid via signed capability-token links (no login), winners are chosen by a pure award algorithm ranking on per-truck rate, with append-only bids, cancellation windows, truck reallocation, delivery-proof trips, and vendor credit scoring. NestJS + Prisma monorepo with a shared Zod/award package.

NestJS
TypeScript
PostgreSQL
Prisma
React
Zod
WhatsApp API

Dristi — Work Summary SMS Automation

A module for Dristi, the app managing the A1 Polymer business (~700 users) at Anwar Group. An n8n workflow generates and sends SMS updates summarizing each responsible person's work, integrated with the app's PHP/MySQL backend to fetch and log work data.

n8n
WhatsApp API
PHP
MySQL

Ulka Chat

Production-ready AI chatbot backend at Saratech Ltd — RAG pipelines and vector search over Qdrant, with MCP-based tool orchestration and scalable FastAPI services.

Python
FastAPI
RAG
MCP
LangChain
LangGraph
Qdrant

Live Demo

FactDrop

Agentic RAG fact-checker where AI agents plan searches, gather multi-source evidence via a custom MCP server, and trust-score every source to produce explainable verdicts — with Bangla–English support.

Next.js
TypeScript
PostgreSQL
Agentic RAG
MCP
LLMs

Live Demo

Memma Logistics Hub

SSR React 19 marketing site for a Bangladesh logistics company on TanStack Start, with a hardened single-table Supabase pipeline (RLS insert-only) for secure quote-request submissions.

TanStack Start
React
TypeScript
Tailwind CSS
Supabase

Live Demo

CampusMate AI

AI assistant for Southeast University with contextual RAG, tool-calling chatbot workflows, and structured knowledge pipelines for accurate, context-aware student support.

Contextual RAG
Python
FastAPI
Pinecone
PostgreSQL

Live Demo

Punji

Flutter finance app for expense tracking, budgeting, and shared expenses with relational expense partners — plus an AI chatbot for money-management assistance.