Sadidul I.

With 6+ years of experience and 100+ completed projects, I specialize in delivering end-to-end AI solutions—from training and fine-tuning models to production deployment. My expertise spans Computer Vision, OCR, LLMs/RAG, and Predictive Modeling, helping businesses automate processes and gain actionable insights. Core Competencies: - Computer Vision & OCR: Object detection (YOLO, ViT), document extraction, and structural drawing analysis using OpenCV, Tesseract, and AWS Textract. - LLMs & Generative AI: Fine-tuning (Llama, GPT), RAG implementation, chatbots, and text-to-speech agents using Hugging Face, ChromaDB, and Pinecone. - MLOps & Deployment: Containerizing services with Docker and deploying scalable solutions on AWS and GCP with a focus on cost optimization and low latency. - Predictive Modeling & Recommendations: Building robust models for fintech, stock forecasting, and personalized recommendation engines. I combine strong Python programming with deep learning frameworks (PyTorch, TensorFlow) to turn complex data into production-ready applications. Let’s discuss how I can help you build or optimize your AI infrastructure. Skills: Python, PyTorch, TensorFlow, Hugging Face/Transformers, RAG, OpenCV, YOLO, ViT, VGG, ResNet, DocTR, Tesseract, AWS Textract, Google OCR, OpenAI API, LLAMA-3, ChromaDB, Pinecone, Docker, AWS, GCP Types of projects I have completed in recent years - #1 OCR / Document Extraction - NLP expert to extract information from natural text — Extract information from images and PDFs using OCR - Computer Vision Expert for Web App (OCR + AutoML) — Testing kit detection from images and reading test results - NLP expert to extract information from natural text — Finding names, company names, etc. from images and PDFs - Computer Vision Engineer for Post-Tension Structural Drawing Analysis (PDFs) — 2D drawing analysis #2 Computer Vision - Object Detection/Counting Project — Iron rod counting in batches using YOLOv8 - Object detection for ice hockey video — Detecting players, puck, and possession based on activity - Mechanical Component Segmentation and Clustering — 2D drawing analysis - AI/ML Engineer, Python/React for Skincare App — ML model integration in React Native app #3 LLM / Generative AI - Prototype Real-Time Text-to-Speech AI — Calling agent using OpenAI (GPT-4, Whisper), Google TTS, and Twilio - Llama-2, Long Llama Engineer — Using open-source LLM models - Data Engineer ML/AI — Customer care chatbot using OpenAI and LLaMA-7B - AI Training for Foreign Language — Fine-tuning LLM models #4 Regression / Predictive Modeling - Cloud-based Data Scientist (AWS) — Models for fintech (NLP, Tabular) - Stock market price prediction — Using historical data to predict future prices - Balance/debt forecasting — Using bank statement data to predict debt and balance #5 Recommendation Systems - Food Recommendation Engine — Correlating users and food menu item based on description - Similar Products based on Users Interest — Find the product a user might be interested in - Perfume suggestion based on User Profile — Suggest existing perfumes based on users profile