About

Machine Learning Engineer with 5+ years of experience designing AI system architectures and building production-ready solutions across vision, language, and multimodal tasks. Specialized in LLMs (GPT-4o, Claude 3, Gemini 1.5), agentic frameworks (CrewAI, LangGraph), RAG with hybrid vector search, and time-series models (Chronos-Mamba, PatchTST). Proficient in end-to-end ML pipelines and scalable deployment using FastAPI, Ray Serve, vLLM, and cloud platforms. Experienced in LLMOps, real-time inference, and integrating cutting-edge AI technologies into intelligent, modular systems.

  • Birthday: 03 October 1997
  • Phone: +8801630830400
  • Phone: +8801915226890
  • City: Dhaka, Bangladesh
  • Email: tanbinirohan@gmail.com

Interests

Data Science

Machine Learning

Computer Vision

Natural Language Processing

Deep Learning

MLOps & LLMOps

Generative AI

Multi-Agent Systems

Education

Notre Dame College

Higher Secondary Certificate

March 2013 - April 2015
Relevant Coursework
  • Math
  • Physics
  • Chemistry
  • ICT

Khulna University of Engineering & Technology

B.Sc. Eng. in Electrical and Electronic Engineering

March 2016 - March 2020
Relevant Coursework
  • Data Structure and Algorithms
  • Digital Signal Processing
  • Digital Image Processing

Jahangirnagar University

Masters in Applied Statistics and Data Science

September 2023 - Present
Relevant Coursework
  • Advanced Statistics
  • Machine Learning
  • Data Analysis
  • Statistical Modeling

Online Certification

Deploy Models with TensorFlow Serving and Flask

Google Cloud Fundamentals: Core Infrastructure

Natural Language Processing in TensorFlow

Improving Deep Neural Networks: Hyperparameter Tuning, Regularization and Optimization

Sequence Models

Experience

TulipTech

April 2025 - Present

AI Engineer

  • Built and deployed next-gen AI inference pipelines using GPT-4o, Claude 3, Gemini 1.5, Mixtral, Llama, and DeepSeek; optimized latency by 30% with vLLM, TensorRT-LLM, LiteLLM, and Ollama on RunPod and AWS Inferentia.
  • Developed multi-agent systems with CrewAI, LangGraph, and AutoGen, enabling dynamic task planning, reflection, and hybrid RAG with ColBERT, BM25, and streaming pipelines via WebSockets.
  • Implemented LLMOps and observability workflows using Weights & Biases, Phoenix, and MLFlow, with robust function-calling, self-consistency, and Mixture-of-Experts (MoE) architectures for scalable agent orchestration.
  • Engineered real-time microservices and APIs using FastAPI, gRPC, Redis Streams, and Kafka; integrated vector databases (Pinecone, ChromaDB, FAISS, Weaviate, LanceDB) ensuring 99.9% uptime for critical GenAI deployments.

SSL Wireless

February 2024 - April 2025

Data Scientist

  • Designed and deployed scalable APIs for edge-based computer vision, improving image classification accuracy by 40% and reducing latency by 20%; developed real-time object detection and tracking models, increasing accuracy by 25% and lowering false positives by 30%.
  • Researched and integrated cutting-edge LLMs (GPT-4, LLaMA, Mistral) with Retrieval-Augmented Generation (RAG), boosting query resolution by 35%; built multimodal systems combining vision and language models to reduce manual effort by 20% and improve decision-making efficiency.
  • Automated finance and business pipelines using Claude, Gemini, and Mixtral-powered LLM agents for forecasting, anomaly detection, and report generation; mentored junior team members on model development and best practices to foster in-house AI expertise.

Easital Technologies Ltd

August 2023 - January 2024

Artificial Intelligence Engineer

  • Led the development of a chatbot for Beeda's international mega-app, leveraging LangChain, Llama Index, and graph databases to manage complex user queries, increasing user engagement by 20%.
  • Reduced OCR errors in business card scanning by 35% through advanced AI-based optical character recognition and image preprocessing techniques; fine-tuned LLMs and engineered prompts to meet organizational and user-specific objectives.
  • Optimized deployment workflows by integrating Docker, Kubernetes, and CI/CD pipelines, resulting in a 25% improvement in deployment speed and reliability.

Aqualink Bangladesh Ltd

May 2021 - June 2023

Machine Learning Engineer

  • Built predictive models for agricultural and weather datasets, improving accuracy by 15% through multimodal feature integration and advanced hyperparameter tuning aligned with business needs.
  • Designed a Django-based real-time analytics platform and visual dashboards using Power BI, Google Data Studio, and Python, boosting operational efficiency by 30% and enabling data-driven decision-making.
  • Collaborated with agronomy and climate experts to deploy models and tools that seamlessly integrated into field operations and supported strategic planning.

L A Technologies

August 2020 - February 2021

Junior Software Engineer

  • Analysis and implementation of deep learning models in non-identical systems using different APIs.
  • Network-related system backend handling.

Intense Ltd.

March 2021 - May 2021

Programmer Analyst

  • Analysis and execution of API documentation of the E-Commerce platform.
  • Integrated SSL commerce with the existing system.

Publications

Projects

  • All
  • Machine Learning
  • Web

Twitter Analysis

End-to-End-Data-Engineering

Network Intrusion Detection System

Plant Disease Detection

OpenCV Projects

NLP-based-English-and-Spanish-Paraphrasing

NLP-based-English-and-Spanish-Paraphrasing

Bangla-Text-Summarizer

Crop Recommendation System

ChatGPT-Chatbot

Core Competencies

Foundation Models & Multimodal LLMs

GPT-4o, Claude 3, Gemini 1.5 Pro, Mixtral, LLaMA 3, Mistral 7B, DeepSeek-VL, Command R+, PaLM 2, LLaVA, Flamingo, Kosmos-2, Whisper, wav2vec 2.0, CLIP, BLIP-2, Segment Anything, SAM-B, Video-LLaMA, Fuyu-Heavy, Idefics2

Time-Series and Forecasting LLMs

Chronos-Mamba, PatchTST, TiDE, TimeGPT, TimesFM, TS-Mamba, Informer, FEDformer, Autoformer, Temporal Fusion Transformers (TFT), DeepAR

Agentic Frameworks

CrewAI, LangGraph, AutoGen, MetaGPT, AgentOps, OpenAgents

LLMOps and Monitoring

Weights & Biases, Helicone, Trulens, OpenLLMetry, PromptLayer, Arize AI, LangSmith, Evidently AI, HumanLoop

Programming Languages

Python, C, C++, SQL, Bash, PowerShell, TypeScript

ML/DL Frameworks

PyTorch, TensorFlow, Hugging Face Transformers, Scikit-learn, OpenCV, Fastai, LightGBM, XGBoost, PyCaret

System Design & MLOps

FastAPI, gRPC, Ray Serve, REST/GraphQL, Kafka, Redis Streams, Kubernetes, Docker, Celery, Apache Airflow, MLflow, Kubeflow, Argo Workflows

RAG & Vector DBs

LangChain, LlamaIndex, ChromaDB, Weaviate, FAISS, Pinecone, LanceDB, Milvus, Hybrid Search (BM25 + Dense), Multi-vector RAG

Cloud & DevOps

AWS (SageMaker, Bedrock, EC2, Lambda), Azure OpenAI, GCP Vertex AI, RunPod, Replicate, Ollama, Hugging Face Hub, Terraform, Ansible

AISecOps & Responsible AI

Adversarial Robustness Toolbox (ART), LLM Guard, Rebuff, Guardrails AI, SecFix, Fairlearn, Snyk, IBM AI Fairness 360

Contact

My Address

House:98, Road:J-5(Extension Road)

South Banasree, Dhaka-1219

Bangladesh

Social Profiles

Email

tanbinirohan@gmail.com

Contact

+8801630830400

+8801915226890