Available for AI/ML roles & research collaborations
Mahmudul Haque
Sakib
Building Intelligent Systems
A portfolio of AI/ML engineering, research-driven experimentation, and production-minded system design.
700+
Problems Solved
3
Research Works
3.68
CGPA
#0 — Who I Am
Engineering Profile
My work spans research-driven AI, end-to-end ML pipelines, and systems designed to move beyond prototypes.
I'm Mahmudul Haque Sakib, a Data Science-focused CSE graduate who builds AI systems that are technically solid, practical to use, and designed with real-world constraints in mind.
My work spans Computer Vision, Natural Language Processing, and RAG-based systems. I have worked on projects involving medical image analysis, medicinal plant classification, sentiment analysis, and document-aware question answering.
I enjoy building complete ML pipelines, from data preparation and model training to evaluation, Docker-based packaging, CI/CD workflows, and AWS deployment. I care about reliability, explainability, and systems that are ready to move beyond prototypes.
Computer VisionNLPRAG SystemsML PipelinesDockerCI/CDAWS DeploymentExplainable AI
AI/ML Research
Working on medical AI, multilingual NLP, and explainable deep learning for practical, research-driven systems.
Systems Builder
Designing end-to-end ML applications, from experimentation and APIs to deployment-ready products.
Problem Solver
Built strong analytical thinking through 700+ problem-solving challenges across competitive platforms.
Teaching Assistant
Supported programming labs and helped students build stronger foundations in problem solving and coding.
#1 — Inspiration
Cinematic Origins of My Curiosity
A small collection of films and programmes that sparked my imagination long before I built systems of my own.
Gattaca
Star Trek: First Contact
Terminator 3: Rise of the Machines
The Thirteenth Floor
The Matrix
Virus
Who Am I
#2 — Tech Stack
Stack & Strengths
The technical foundation behind my workflows in machine learning, retrieval systems, and product-oriented engineering.
01
Languages
PythonCC++JavaSQL
02
Frontend
ReactNext.jsTailwind CSS
03
Applied AI & LLMs
PyTorchTensorFlowscikit-learnHugging Face TransformersLangChainLangGraphGroqYOLOv10
04
Vector DB / RAG
PineconeChromaDBFAISSRAGLLM Fine-tuning
05
DevOps & Deployment
FastAPIGitDockerGitHub ActionsAWS ECRVercelStreamlitFlaskGradio
#3 — Projects
Built Systems
Projects that reflect how I design, build, and deploy AI systems with practical use cases and technical depth.
01 / Flagship ProjectNLP / LLMsFeatured
MediBot — Source-Grounded Medical QA
A portfolio-ready medical RAG assistant that answers questions from textbook content using Pinecone retrieval, Groq-powered generation, and a Streamlit chat interface.
Why it matters
Medical QA systems can hallucinate when they answer without evidence. MediBot reduces that risk by grounding responses in retrieved textbook chunks and applying safety-oriented fallback behavior.
Tech Stack
PythonLangChainPineconeGroqStreamlitDockerAWS
Project Highlights
- RAG pipeline powered by Pinecone, LangChain, and Groq
- Medical book-only answering with source-grounded retrieval
- Similarity-threshold fallback for low-confidence queries
- Conversation-aware Streamlit chat UX with source display
My Contribution
Built the end-to-end applied GenAI pipeline, including ingestion, retrieval, response guardrails, chat UX, and deployment-ready structure.
02NLP / LLMsFeatured
TaskBot SQL Agent
An AI-powered task management assistant that lets users manage a SQLite database through natural language instead of writing SQL manually.
Tech Stack
PythonLangChainLangGraphSQLiteStreamlitGroq
03NLP / LLMsFeatured
RAG Document Search
An agentic document-question-answering system that ingests files, indexes them in FAISS, and returns grounded answers through a Streamlit interface.
Tech Stack
PythonLangGraphLangChainFAISSStreamlitGroqHugging Face
04NLP / LLMs
Travel Planner Using LangGraph and FastAPI
A graph-based multi-agent travel planning system that combines smart trip parsing with real-world APIs for weather, country data, currency, and distance-aware planning.
Tech Stack
PythonLangGraphFastAPIStreamlitGroqREST APIsCaching
05Computer Vision
Lemon Leaf Disease Detection — Ensemble CNN + XAI
An end-to-end plant disease classification project using multiple CNN backbones, weighted ensembling, and Grad-CAM/Ablation-CAM for interpretability.
Tech Stack
PyTorchTorchvisionResNet50EfficientNetDenseNetGrad-CAMAblation-CAM
06NLP / LLMs
QuickAnswer Serper Agent
A streaming AI web-search assistant that combines a Groq-hosted LLM, live Google Serper search, LangGraph memory, and a clean Streamlit chat interface.
Tech Stack
PythonStreamlitLangChainLangGraphGroqGoogle Serper API
More Projects
Other Work
07NLP / LLMsAdditional Project
Bangla Emotion Classification with Transformers + XAI
A multilingual NLP project for Bangla emotion classification using transformer fine-tuning and SHAP-based interpretability.
Tech Stack
PythonTransformersmBERTXLM-RoBERTaSHAPHugging Face
08NLP / LLMsAdditional Project
Market Assistant Bot
A finance-oriented assistant for Telegram that delivers educational investing guidance and safe market commentary through a FastAPI + Groq backend.
Tech Stack
PythonFastAPIGroqTelegram Bot APIPrompt EngineeringSafety Guardrails
09NLP / LLMsAdditional Project
LangChain Web Search
A conversational AI search assistant that augments a Groq-powered LLM with DuckDuckGo, Arxiv, and Wikipedia tools inside a Streamlit interface.
Tech Stack
PythonStreamlitLangChainGroqDuckDuckGoArxiv APIWikipedia API
#4 — Research
Publications & Ongoing Research
Papers and ongoing work that turn technical questions into structured investigation and measurable contribution.
Accepted — ICIDA 2025International Conference on Intelligent Data Analysis (ICIDA 2025)
KidneyStoneNet: A Deep Learning Approach for Kidney Stone Detection with Transfer Learning and Explainable AI
Mahmudul Haque Sakib · et al.
Abstract
This paper presents KidneyStoneNet, a CNN-based transfer learning framework for automated kidney stone detection in medical images. We train and evaluate ResNet, DenseNet, and NASNet architectures using PyTorch, integrate explainable AI (XAI) techniques for prediction interpretability, and deploy an interactive Gradio web application for real-time inference. Our approach demonstrates that transfer learning models can provide clinically interpretable, high-performance detection with minimal training data.
PyTorchTransfer LearningXAIGrad-CAMGradio
In Preparation
MediLeafNET: An AI-Driven Multimodal System for Medicinal Plant Discovery and Ethnobotanical Knowledge Retrieval in Bangladesh
Mahmudul Haque Sakib · et al.
Vision TransformerBERTLLaMAXAIGradioMultimodal AI
In Preparation
Aspect-Based Sentiment Analysis for Smart Devices: A Transformer-Based Approach
Mahmudul Haque Sakib · et al.
BERTRoBERTaSentiment AnalysisXAITransformersGradio
#5 — Work History
Experience
Roles and responsibilities that shaped how I teach, collaborate, and build with technical clarity.
Undergraduate Teaching Assistant
East West University
March 2025 – June 2025Dhaka, BangladeshPart-time
- ▸Conducted programming lab sessions covering MS Office, HTML, and foundational computing for 50+ undergraduate students.
- ▸Supported debugging and code review, improving student performance in practical assessments.
- ▸Maintained academic documentation and coordinated closely with faculty to ensure seamless course delivery.
- ▸Assisted students with technical questions and provided structured feedback to improve learning outcomes.
Actively seeking AI/ML Engineer, Data Scientist, or Research roles — open to internships and full-time positions.
#6 — Academic Background
Where I Built My Foundation
Educational milestones that laid the groundwork for my technical expertise, research capabilities, and passion for AI.
Feb 2022 – Dec 2025Dhaka, Bangladesh
Institution
East West University
Qualification
B.Sc. in Computer Science and Engineering
Major: Data Science
CGPA
3.68 / 4.00
Academic Summary
Specialized in Machine Learning, NLP, Computer Vision, and AI system design.
Merit Scholarship recipient (Spring 2025) and Undergraduate Teaching Assistant (Spring 2025), with relevant coursework in Deep Learning, Natural Language Processing, Computer Vision, Data Mining, and Software Engineering.
Specialization & Research Interest
Artificial IntelligenceMachine LearningDeep LearningData SciencePythonResearchInnovation
2020Narsingdi, Bangladesh
Institution
Narsingdi Model College
Qualification
Higher Secondary Certificate (HSC)
Major: Science
GPA
4.83 / 5.00
Academic Summary
Completed higher secondary studies in the Science stream with strong academic performance.
Focus Areas
MathPhysicsChemistry
2018Narsingdi, Bangladesh
Institution
Bhuim High School
Qualification
Secondary School Certificate (SSC)
Major: Science
GPA
4.78 / 5.00
Academic Summary
Developed early analytical skills through core science subjects.
Focus Areas
MathPhysicsChemistry
#7 — Achievements
Milestones That Matter
A record of competitive, academic, and research recognition earned through consistency and performance.
2025
Scholarship
Merit Scholarship
East West University
Awarded for maintaining a GPA of 3.90+ across three consecutive semesters — recognizing sustained academic excellence.
2024
Scholarship
Medha Lalon Scholarship
East West University
Awarded for maintaining a GPA of 3.50+ across three consecutive semesters in Computer Science & Engineering.
2024
Competition
Pupil Rank — Codeforces
Codeforces
Achieved Pupil rank (1255 rating) on Codeforces, solving 700+ algorithmic problems across competitive programming platforms.
View Profile2024
Competition
2★ Rating — CodeChef
CodeChef
Earned a 2-star rating (1535) on CodeChef through consistent competitive programming performance.
View Profile2025
Award
Research Accepted — ICIDA 2025
ICIDA 2025 International Conference
First-authored research on kidney stone detection with explainable AI accepted at the International Conference on Intelligent Data Analysis (ICIDA 2025).
#8 — Learning & Credentials
Certifications
Verified certifications and learning milestones that support my practical AI/ML, agentic systems, and production-oriented engineering skill set.
CertifiedApr 12, 2026
AI Engineer for Data Scientists Associate
DataCamp
Associate-level certification focused on applying AI engineering concepts in practical data science workflows.
AI EngineeringData ScienceModel DevelopmentApplied AI
ID: AEDS0016556238778
CertifiedApr 5, 2026
RAG for Generative AI Applications
IBM
Specialization focused on building GenAI applications with Retrieval-Augmented Generation (RAG), covering vector search, retrievers, and end-to-end RAG workflows.
RAGGenerative AIVector DatabasesRetrievers+3
ID: ZEW0KORZ0KJUVerify
CertifiedMar 26, 2026
Advanced RAG with Vector Databases and Retrievers
IBM
Advanced RAG course focused on retrievers, vector database workflows, and stronger retrieval pipelines.
Advanced RAGRetrieversVector DatabasesLLM Systems
ID: YJLWLHM748G7Verify
CertifiedApr 8, 2026
Agentic AI with LangChain and LangGraph
IBM
Hands-on course focused on building agentic AI systems using LangChain and LangGraph.
Agentic AILangChainLangGraphWorkflow Orchestration
ID: J4VQB6KR0OUVVerify
CertifiedApr 11, 2026
Build AI Agents using MCP
IBM
IBM course focused on building AI agents using MCP and practical agent communication patterns.
AI AgentsMCPAgent CommunicationIBM Skills Network
ID: BAOOEZQEGL60Verify
CertifiedApr 7, 2026
Fundamentals of Building AI Agents
IBM
Foundational course covering the core concepts, patterns, and workflows behind AI agent development.
AI AgentsAgent WorkflowsLLM SystemsIBM Skills Network
ID: QQDLCWHO57SPVerify
#9 — Opportunities
The Roles I’m Ready For
The kinds of opportunities that align with my skills, research interests, and long-term goals.
AI/ML Engineering
Roles focused on building, training, and deploying machine learning models — especially in NLP, computer vision, and generative AI. Comfortable working across the full ML lifecycle from data to production.
LLMsComputer VisionNLPMLOpsRAG
Research-Oriented Roles
Positions at the intersection of research and engineering — contributing to AI publications, building experimental systems, and solving open problems in medical AI, multilingual NLP, or explainable AI.
XAIMedical AIMultilingual NLPBenchmarking
Data Science & Analytics
Data-heavy roles where I can apply statistical modeling, feature engineering, and visualization to extract insight from complex datasets. Interested in combining domain knowledge with modern ML pipelines.
Data AnalysisFeature EngineeringPower BIPredictive Modeling
Open to full-time roles, internships, and research collaborations — locally in Dhaka or remotely worldwide.
#10 — Contact
Let’s Build Together
Whether it’s a team, a project, or a research problem, I’m open to meaningful technical collaboration.
Let's build something meaningful.
Reach out for AI/ML roles, collaborations, research, or project discussions.
Contact Channels
Let's connect professionally.
GitHub
@SAKIB0004LinkedIn
mhsakib0004Gmail
mahmudulhaque540@gmail.comCodeforces
Pupil · 1255CodeChef
2★ · 1535K
Kaggle
mhsakib4