I'm a Senior Data & AI Engineer with 7+ years of experience architecting scalable data pipelines, MLOps infrastructure, and production-grade AI applications. Currently building AI for security surveillance at NewSense Eng. in Munich and finalizing my M.Sc. in Data Science at FAU Erlangen-Nuremberg.
- 🔭 Currently working on 3D perception models and RAG-based LLM applications for enterprise use
- 🌱 Specializing in Generative AI, RAG architectures (LangChain, FAISS, Pinecone), and model evaluation (LangSmith / RAGAS)
- 🏆 1st Place — Adidas Data Science League, Herzogenaurach (2024)
- 🚀 Founder of iPractest.com — an AI-powered language learning platform
- 📚 Published in Elsevier — Information Sciences on cluster-oriented instance selection
- 🗣️ Languages: English, Hindi, Urdu (Fluent) · German (A2) · Bangla (Native)
Languages
ML / AI / LLMs
Data Engineering & MLOps
Cloud, Infra & Backend
AI Application Engineer — NewSense Eng., Munich · Sep 2025 – Present Building AI security surveillance applications with Angular + FastAPI + gRPC. Improving 3D perception models for computer vision and developing RAG-based LLM apps with LangChain, FAISS, and Pinecone.
Lead Data & AI Engineer — Mi-C3 Int. Ltd., Malta · Feb 2023 – Aug 2025 Designed real-time ETL pipelines with PySpark integrated into Snowflake. Deployed orchestrated workflows on Docker + Kubernetes. Built monitoring stacks with Grafana, Superset, and Pub/Sub. Shipped an AI chat agent for client information retrieval using LLM + RAG + LangChain.
Data Engineer / Software Engineer — SincoS Automation Technologies, Dhaka · Nov 2019 – Feb 2023 Built IoT data pipelines with AWS Glue, trained anomaly-detection models on SageMaker, and developed full-stack Spring Boot + React/Angular applications including cross-platform mobile apps in Flutter.
| Project | Description | Stack |
|---|---|---|
| RAG Document Assistant | LLM retrieval system for enterprise documents | LangChain · FAISS · OpenAI API |
| Anomaly Detection | Production anomaly detection with retraining loop | XGBoost · SageMaker · Kubernetes · MLflow |
| Soap Hardness Prediction | Predictive ML pipeline with automated retraining | Databricks · MLflow · XGBoost |
| Energy Monitoring System | Real-time IoT monitoring with ML insights | Airflow · DBT · RabbitMQ · Angular |
| Telecom Tower Data Processing | Streaming pipeline for telecom infrastructure | NiFi · RabbitMQ · PostgreSQL · AWS |
| iPractest.com | AI-powered language learning platform | LangChain · Hugging Face · React · Flask |
- Cluster-oriented instance selection for classification problems — Elsevier, Information Sciences (2022)
- Solving multi-class classification tasks with classifier ensemble based on clustering — UIU Institutional Repository (2019)
- 🟢 NVIDIA Certified Professional: Agentic AI
- 🥇 1st Place — Adidas Data Science League (Herzogenaurach, 2024)
- 📘 Machine Learning with Python — Coursera
- ⛓️ Certified Blockchain Developer™ — Blockchain Council
💡 Open to collaborations on Generative AI, RAG systems, and scalable data infrastructure.