My goal is to work as an Applied AI Engineer on projects where I can contribute and learn, with a strong backend and DevOps orientation focused on building scalable production-grade AI systems. I am passionate about building automated and maintainable AI-driven systems that integrate Generative AI and cloud-native DevOps practices.
I am currently working at Mediusware Ltd. as a Software Engineer(I) in AI/ML domain. I build and improve backend-driven AI systems for real-world applications.I have hands-on experience working on both standalone AI applications and integrating AI-based features into existing software systems. My work focuses on designing RAG pipelines, multi-agent workflows, scalable APIs, and AI-powered backend services using technologies such as FastAPI, LangChain, LangGraph, vector databases, and modern LLM frameworks. I have completed my BSc in Information Technology from the Institute of Information Technology, Jahangirnagar University. Through my professional work, research involvement, and mentorship activities, I have developed practical experience in Machine Learning, Deep Learning, Natural Language Processing, and Generative AI. I am particularly interested in building maintainable and production-ready AI systems that combine back.
Frameworks -- Python and Generative AI
Flask,FastAPI,Django,Streamlitresearching, building, and designing the artificial intelligence responsible for machine learning and maintaining and improving existing artificial intelligence systems.
Creating Documentation specializing on ML , DL and GenAI concepts and continue my research work comparative analysis of LLM models
Design, build, and maintain data pipelines for collecting, processing, and storing large datasets for fine-tuning the Large Language Model.
Collecting data for a research paper "Sentiment Analysis on Bengali Social Media Comments."
CGPA : 3.47.
Designed and deployed a multi-agent system to automate workflows in the vehicle dealership domain via chatbot. Built agents for inventory extraction, FAQ answering (RAG), trade-in price estimation, call scheduling, and conversation closure.
Technologies: LangGraph, LangChain, RAG, Hugging Face
Developed a chatbot capable of handling multimodal queries (text + image + voice) across structured and unstructured data. Integrated speech-to-text and TTS, and implemented re-ranking models for enhanced response accuracy.
Technologies: Hugging Face (Gemma, Whisper, Kokoro), LangChain, RAG
Created a personality-aware habit tracking application that identifies user traits through guided questions. Implemented a 21-day cycle with reward-based motivation and deployed as a scalable AI service.
Technologies: OpenAI (gpt-4o-mini), FastAPI, Render
Built a recursive web scraping and ETL pipeline for unstructured data. Automated data scraping, cleaning, and structuring into knowledge-ready formats, enabling future chatbot and AI integration.
Technologies: Crawl4AI, FastAPI
I like to experience updated frameworks and libraries & uncover hidden data insights also explore how can I implement them in existing projects for better performance
Below are the online courses I have completed.
| Course Name | Platform/Provider | Certificate Link |
|---|---|---|
| Generative AI Mastery | Udemy | View Certificate |
| Microsoft Excel | 10 Minute School | View Certificate |
| Python Beginners to Advanced Bootcamp | Udemy | View Certificate |
| Prompt Compression and Query Optimization | DeepLearning.AI | View Completion |
| Orchestrating Workflows for GenAI Applications | DeepLearning.AI | View Completion |
| ChatGPT Prompt Engineering for Developers | DeepLearning.AI | View Completion |
| Knowledge Graphs for RAG | DeepLearning.AI | View Completion |
Position : 2nd Runner Up
INNOVATION SHOWCASING : Smart Jahangirnagar University
Below are the details to reach out to me!