I turn unstructured documents, complex models, and ambitious product ideas into
fast, dependable, production-ready AI applications people can actually use.
Currently working as a Mid-Level AI Engineer at SparkTech Agency, where I own
cross-functional teams and build scalable backend systems with FastAPI, RAG, OCR, NLP, computer vision, automation, and MLOps systems.
|
AI-powered U.S. route-processing platform that extracts route information automatically from permit documents using AWS Textract, OpenAI, NLP, and FastAPI. Contribution: System architecture and end-to-end AI route-processing pipeline. |
Production-grade legal assistant for U.S. Federal and Michigan State law, powered by semantic retrieval, article-level indexing, and LLM-based reasoning. Performance: Optimized retrieval and orchestration for sub-2-second responses. |
|
Muslim AI - Get instant answers to your spiritual questions with integrated AI, trained to provide context-aware guidance from reliable sources. Role - I was the R&D lead of that project, did the System feasibility test, and also architected the cloud system. |
Production-grade AI-powered intake and workflow automation platform that processes requests from WhatsApp and Outlook, analyzes text and images, detects missing information, classifies risk, and intelligently routes requests for automated or human-assisted action. |
|
Currently in Apple Apstore Testflight Mode
Scalable event-driven backend for an AI-powered fitness storybook platform using FastAPI, PostgreSQL, Redis, Docker, ECS, SQS, S3, Bedrock, and EC2. Heavy AI workloads were decoupled from the main API using an event-driven microservices architecture to eliminate timeouts and improve system reliability. |
|
Current Role : Mid-Level AI Engineer
Timeline : September 2025 – Present
Career Path :
Junior AI Engineer : September 2025 – June 2026
Mid-Level AI Engineer : July 2026 – Present- Promoted to own a cross-functional team of 3–4 engineers, overseeing system design, client communication, technical delivery, and engineering coordination.
- Successfully delivered 12+ production AI systems spanning RAG, OCR, NLP, computer vision, and document intelligence.
- Engineered scalable RAG pipelines using LangChain, FAISS, Pinecone, FastAPI, and LLM orchestration, achieving sub-2-second latency.
- Established automated MLOps workflows with Weights & Biases, AWS, Docker, and CI/CD, accelerating model experimentation and production deployment.
- Worked across multiple responsibilities as an AI Engineer, R&D Lead, Python Developer, and System Architect.
Position : Software Engineer
Timeline : August 2024 – October 2025
Rover : CLADOS- Developed and programmed hardware systems for CLADOS, the GUB Mars Rover Team’s primary rover.
- Contributed to the rover that ranked 6th worldwide and 1st in Asia at the 2026 Anatolian Rover Challenge.
- Collaborated across software and hardware teams to integrate and test rover subsystems.
Degree : Bachelor of Science in Computer Science & Engineering
University : Green University of Bangladesh
Timeline : February 2022 – January 2026
Major : Artificial Intelligence and Machine Learning
Focus Areas : Generative AI · NLP · Computer Vision · MLOps · XAI|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
NeuroGebra is an open-source symbolic-neural framework for visualizing machine-learning mathematics and neural-network operations in real time.
- Makes model formulas, training behavior, gradients, and neural operations easier to understand.
- Works as a learning companion alongside PyTorch and TensorFlow.
- Adopted by 8,000+ students and researchers worldwide.
- Published on PyPI with public documentation and a permanent DOI.
pip install neurogebraAwards:
- Result : 1st Runner-Up
Competition : Comilla University Data Science Hackathon
Date : November 2024
- Result : Best Innovation Award
Competition : IDP-1 University Project Showcase
Project : ViolaShield
Date : January 2025
- Result : Finalist
Competition : NSU FutureBuilders AI Hackathon
Date : December 2024
- Result : Finalist
Competition : BUBT AI Competition
Date : November 2025
Leadership:
- Role : Information Secretary
Organization : GUB Computer Club
Timeline : 2023 – 2024
Impact : Organized 5+ ML-focused workshops for 200+ students
- Role : Enduring Member
Organization : BASIS Students' Forum
Impact : Mentoring members on industry practices and career developmentI’m open to AI engineering roles, research collaborations, open-source work,
and opportunities to build ambitious AI products.
