I'm Md Hasan Imon an experienced AI/ML Engineer from Bangladesh, specializing in Machine Learning, NLP, Deep Learning, Generative AI, LLM Fine-Tuning (LoRA/QLoRA), Agentic RAG Pipelines, and Vector Databases. Over multiple production-grade projects, I've designed and deployed end-to-end, multi-agent, tool-augmented AI systems that deliver measurable business outcomes through reasoning, automation, and intelligence at scale.
With strong proficiency in FastAPI, Flask, Docker, CI/CD with Backend-as-a-Service (BaaS), I ensure every system is enterprise-ready, modular, and deployable across real-world environments. My work bridges deep technical design with business impact — enabling organizations to harness AI for smarter decisions, scalable automation, and real competitive advantage.
Designing end-to-end AI/ML pipelines from data to deployable applications.
Crafting modular agentic workflows with memory-integrated generative AI architectures.
Fine-tuning LLMs using LoRA, QLoRA, PEFT, and quantization techniques.
Enabling intelligent decision-making via tools, memory, and multi-agent planning.
Building intelligent systems using structured, labeled, and unlabeled datasets efficiently.
Deploying scalable ML models using Streamlit, FastAPI, and local servers.
I specialize in building full-stack, real-world AI/ML systems using cutting-edge technologies such as Agentic AI, LangGraph, LLM Orchestration, and Multi-Agent Architectures. My work focuses on intelligent planning, reasoning, and tool-augmented AI assistants.
Pursuing BSc in Computer Science and Engineering from City University, Dhaka, Bangladesh. Specialized in Software engineering principles with a focus on real-world applications.
Actively building industry-grade AI systems, participating in research, and contributing to open-source AI/ML projects during academic years (2022 – 2025).
During this phase, I focused entirely on academic excellence in science subjects including Physics, Chemistry, Mathematics, and Biology. While I did not engage in extracurricular activities at this stage, this period laid the foundational knowledge and discipline that later shaped my journey into AI/ML engineering.
Successfully completed the Secondary School Certificate under the Dinajpur Board, laying the academic foundation in core subjects. This stage served as the starting point of my educational journey.
Designed, developed, and deployed backend-centric AI/ML systems using Python, integrating models into production-grade services and APIs. Built and maintained scalable, modular AI backends including RAG and multi-agent workflows. Implemented end-to-end data and inference pipelines with optimized preprocessing, model serving, and performance-focused backend execution.
Worked as a Backend AI Engineer focusing on building production-grade multi-agent AI systems. Developed and maintained agentic workflows using LangGraph and AgentOps orchestration. Implemented tool-augmented pipelines (PDFs, Web, APIs, Databases) and integrated memory-driven reasoning systems. Fine-tuned large language models (LLMs) using LoRA, QLoRA, and PEFT for real-world AI applications such as intelligent financial advisors and medical assistants.
How I designed a multi-agent system (Planner, Tool, Memory) with LangGraph to analyze portfolios, fetch market data via tools, and deliver explainable, goal-aware recommendations.
Inside the architecture: symptom triage agent, retrieval-augmented grounding on medical guidelines, citation fidelity checks, and fallback policies to maintain safety while answering patient queries.
A walkthrough of my pipeline for claim parsing, Fine Tuning, reranking, and structured evidence scoring—plus how I log traces and evaluate reliability with AgentOps.
Savar, Dhaka, Bangladesh