Sourav Sarkar is an AI Engineer who holds an MTech in CSE with a specialization in AI and Data Science from IIIT Kota. He has hands-on experience developing advanced applications using LLMs (Mistral, Langchain) and computer vision techniques (YOLO V5). His expertise includes RAG implementation, document classification, and working with deep learning frameworks like PyTorch and TensorFlow.
Experience
AI Engineer
WESEE, Indian Navy
Built Document Question Answering Application using Mistral LLM, Langchain and Chroma DB with Retrieval Augmented Generation (RAG). Instruction Tuned Mistral 7B for document classification, extracting text from PDF, PPTX and DOCX formats. Developed an application for Warship Detection and classification from Large Satellite TIF images using YOLO V5 with Oriented Bounding Boxes. Extensively worked on Audio Denoising, Speaker Recognition, Translation and Transcription. Exposed and integrated with Huggingface open source models.
Artificial Intelligence Intern
WESEE, Indian Navy
Fine tuning of Bert and GPT-2 to compare their performance on synthetic text classification. Mask RCNN Implementation for Warship Detection from Satellite Imagery.
Data Science Intern
Jaipur Modern
Time Series Analysis, Restaurant Sales Data Preprocessing, Restaurant Sales Forecasting.
Education
Indian Institute of Information Technolgy, Kota
M.Tech CSE
AI and Data Science
Amity University Kolkata
B.Tech CSE
CSE
Licenses & Certifications
Neural Network from Scratch in TensorFlow
Coursera Project Network
Exploratory Data Analysis with Seaborn
Coursera Project Network
Crash Course on Python