Mehedi Hasan

Full-Stack Engineer & AI Researcher

|

I love exploring the under-the-hoods of technologies and CS fundamentals. I'm working with Model Context Protocol (MCP), LangChain, RAG systems, FastAPI, Django, Node.js, React, and Next.js to build scalable AI-driven platforms.

About

Hi, I'm a software engineer and competitive programmer with a deep love for CS fundamentals. I enjoy exploring under-the-hoods of complex tech to understand its core

Currently a Software Engineer Intern (AI/ML) at Samsung R&D Institute Bangladesh, building AI agents on internal user data and documents. I have built AI systems including RAG pipelines, an LLM-driven data manager powered by Model Context Protocol (MCP), a dubbing platform and developed web solutions to streamline operations and strengthen digital infrastructure.

Location

Sylhet, Bangladesh

Focus Areas

Full-Stack
RAG Systems
LLM Integration

Technical Expertise

Full-Stack Development

End-to-end application development from database design to frontend implementation

AI/ML Integration

RAG pipeline design, vector database optimization, semantic search, LLM orchestration

API Design & Integration

RESTful APIs, third-party service integration

Performance Optimization

Database query optimization, caching strategies

System Architecture

Scalable database schemas, containerization

DevOps & Deployment

Docker containerization, Cloud deployment

Tech Stack

Languages

C/C++
Python
JavaScript/TypeScript
Java

Web Technologies

Node.js/Express
React
Next.js
Django
FastAPI
Tailwind CSS

AI/ML Integration

RAG
Vector Databases
LangChain
LLM APIs
MCP

Databases

MySQL
MongoDB

Cloud & DevOps

AWS
Docker
Linux

Competitive Programming

2000+
Problems Solved
3
National IUPCs

DSA Knowledge

Data Structures

Arrays & Linked Lists
Stacks & Queues
Hash Tables
Heaps & Priority Queues

Trees

Tree Algorithms
Segment Trees
Fenwick Trees
Trie & Suffix Structures

Graph Algorithms

BFS & DFS
Dijkstra / Bellman-Ford
MST (Kruskal, Prim)
Topological Sort

Paradigms

Dynamic Programming
Greedy Algorithms
Divide & Conquer

Analysis

Time & Space Complexity
Optimization Techniques

Projects

AI Document Companion
Chat with your documents — PDFs, Word, PowerPoint, Excel, Markdown, CSV, JSON, and more. I built this to understand how these tools actually work: Unstructured for layout-aware parsing, LangChain LCEL for composing the RAG pipeline, FastAPI's async engine for real-time streaming via SSE, and Ollama to run everything locally with no external API calls. Answers stream token by token with exact source citations. Fully offline.
FastAPI
LangChain
ChromaDB
+4
AI Data Brain
AI Data Brain
Manage any data by just talking to it. Type or speak in Bengali or English — the system uses MCP tools and LangChain to turn that into real database operations: insert, query, update. MCP handles schema discovery and validation so it works on any table without hardcoding. Built on Next.js frontend and Django backend with WebSocket for real-time responses.
MCP
LangChain
Next.js
+4
Buddy Script
Built as a selection task for Appifylab's Full Stack Engineer role. A social media feed where users create public/private posts with images, interact through nested comments and replies, and like/unlike at every level. JWT + refresh token auth, cursor-based pagination, Redis rate limiting, and Cloudinary for image uploads.
Next.js
React
TypeScript
+5
ClearVocals.ai - AI Dubbing Platform
ClearVocals.ai - AI Dubbing Platform
Full-stack AI dubbing and audio separation platform integrating AssemblyAI and DIA Voice Cloning APIs for high-fidelity audio processing and voice synthesis.
Next.js
TypeScript
Express.js
+3
Brikkho Bondhon
Brikkho Bondhon
A community-driven platform dedicated to environmental conservation through tree planting. Features include volunteer mobilization, donation management for reforestation projects, and global community engagement to combat climate change.
Next.js
Node.js
CitationAI
CitationAI
Retrieval-Augmented Generation (RAG) system for academic research papers, featuring hierarchical, section-aware document chunking and intelligent citation tracking. Designed and implemented a hybrid retrieval pipeline to improve relevance and contextual accuracy. Integrated Redis-based caching and optimized query execution to ensure responsive and efficient research workflows.
FastAPI
PostgreSQL
Qdrant
+4
Game Hub
Game Hub
A video game discovery platform built with React and the RAWG API, featuring real-time search, multi-criteria filtering, and a responsive Chakra UI design with skeleton loading and dark mode.
React
TypeScript
Vite
+1
IoT-Based Home Control System
A smart automation system enabling remote control of home appliances from anywhere in the world. Built using the ESP8266 microcontroller for Wi-Fi connectivity and the Blynk platform for a seamless mobile interface.
ESP8266
Blynk
C++
WordPress Plugin Generator
WordPress Plugin Generator
Automated WordPress plugin generation system using LangChain, RAG pipelines, custom MCP servers, and multi-LLM orchestration (Claude, Gemini, OpenAI) for intelligent plugin development.
LangChain
RAG
MCP Servers
+4

Experience & Journey

Software Engineer Intern (AI/ML)

work

Samsung R&D Institute Bangladesh

Apr 2026 - Present
Dhaka, Bangladesh

Building AI agents on internal user data of Samsung apps and documents. Working on RAG pipelines and agentic systems that let internal teams query and interact with proprietary data.

Trainee Software Engineer

work

recurse.ai

Feb 2025 - Jul 2025
Remote

Designed and developed full-stack AI dubbing platform (clearvocals.ai) with Next.js, Express.js, integrating AssemblyAI & Fish-speech open source Voice Cloning. Built automated WordPress plugin generation system using LangChain, RAG pipelines, custom MCP servers, and multi-LLM orchestration (Claude, Gemini, OpenAI). Developed production-ready WordPress assistant chatbot with RAG, real-time web search, and autonomous content generation.

Lead Developer

work

Brikkhobondhon IT Department

2025 - Present
Bangladesh

Leading the Office Digitalisation Project, developing website and mobile app solutions to streamline organizational operations and enhance digital infrastructure.

B.Sc. Computer Science & Engineering

education

Shahjalal University of Science and Technology (SUST)

2022 - 2026 (Expected)
Sylhet, Bangladesh

Final year Computer Science student with CGPA 3.61/4.00. Focus on AI/ML, RAG systems, Full-Stack Development, and Competitive Programming. Located in Sylhet, Bangladesh.

Certifications

AWS Academy Graduate - Cloud Foundations
Jan 2026

AWS Academy Graduate - Cloud Foundations

AWS Academy

ID: b1cf8ac4-1a29-4e94-a4cc-2fa5a90ac5fe

AWS
Cloud Foundations
EC2
S3
+2
View Credential
React 18 for Beginners
Apr 2024

React 18 for Beginners

Mosh Academy

ID: cert_wbw1xich

React
Hooks
State Management
Component Design
+1
Certificate of Participation - MIAKE KUET IUPC Onsite 2025
Jan 2025

Certificate of Participation - MIAKE KUET IUPC Onsite 2025

IUPC

Competitive Programming
Algorithms
Problem Solving
Team Collaboration
Certificate of Appreciation - API AVENGERS Finalist
Jan 2025

Certificate of Appreciation - API AVENGERS Finalist

Heckathon

API Development
Backend Engineering
System Design
Competition Excellence

Research

Optimizing Post-Quantum LWE Encryption for Large Image Data

B.Sc. Thesis, SUST(In Progress)2025-26

A comprehensive hybrid cryptographic preprocessing and acceleration framework integrating image compression techniques (RLE, Huffman coding, YCbCr with chroma subsampling) with Strassen's matrix multiplication for LWE-based encryption. Achieved 41.19% average data volume reduction while maintaining lossless compression (PSNR = ∞) and full post-quantum security guarantees.

Post-Quantum Cryptography
LWE
Image Compression
Strassen Algorithm
Lattice Cryptography

BanglaNoise-10: A Diverse Audio Dataset for Machine Learning on Urban Environmental Noise in Bangladesh

Data in Brief (Under Review)2025

First nationwide South Asian environmental noise dataset containing 5,035 labeled audio clips across 10 urban noise categories. Achieved 98% accuracy with Whisper-Base, 97% with Wav2Vec2-Base. Dataset includes bike, bus, car, CNG auto-rickshaw, construction noise, protest, siren, traffic jam, train, and truck sounds collected across Bangladesh.

Audio Dataset
Machine Learning
Urban Sound Classification
Wav2Vec2
Whisper
Deep Learning

Voluntary Works

Writing blogs on Computer Science Fundamentals

work

Knowledge-sharing Platform - mdmehedi.tech

2024 - Present

Created and maintain a knowledge-sharing platform dedicated to computer science fundamentals. Writing in-depth technical blogs exploring core CS concepts, algorithms, data structures, and system design principles. Focus on making complex topics accessible through clear explanations and practical examples.

Coordinator

work

SUST Competitive Programming Club

2024 - 2025
Sylhet, Bangladesh

Managed and conducted training sessions, problem-solving classes, and programming contests to guide and mentor junior students in improving their competitive programming skills. Organized practice rounds, code review sessions, and strategy discussions to help develop problem-solving expertise.

Founding Member

work

Institute of Nebula for Changing

2018 - 2022
Chandpur, Bangladesh

Co-founded an organization dedicated to rural education and empowerment, focused on delivering quality education in underserved rural areas. Designed and delivered training programs in basic computer knowledge and programming to empower rural youth. Contributed to bridging the digital divide and creating opportunities for skill development in remote communities.

Get In Touch

I'm always open to discussing AI projects, collaborations, or engineering challenges. Feel free to reach out!