Headline

AI/ML Engineer | Python Developer | GenAI, RAG & Computer Vision | FastAPI • LangChain • OpenCV | Open to Full-Time Roles 2026

About

Most AI/ML projects die in Jupyter notebooks. Mine went into production at Tata Motors — and I've kept building since. I'm Sarthak — a final-year Computer Science student who doesn't just build models, I ship systems that replace manual work with automation. THE PROBLEM I SOLVED AT TATA MOTORS: Trainee assessments were 100% paper-based — 1,000+ records, manual entry, zero analytics. I built an OCR + ML pipeline (OpenCV + Tesseract + FastAPI, Dockerized) that digitized the entire workflow, improved extraction accuracy by 25%, and moved reporting for 10+ stakeholders from weekly Excel updates to real-time dashboards — cutting evaluation cycle time by 70-80%. WHAT I'VE BUILT SINCE: A RAG-based document Q&A system (LangChain + Gemini + ChromaDB) that lets users query PDFs in natural language with page-level source citations A no-code analytics platform (Streamlit + Scikit-learn) that automates full data science workflows for non-technical users An NLP + clustering pipeline segmenting 10,000+ financial records into trader behavior archetypes WHAT I ACTUALLY DO: GenAI & NLP: RAG pipelines, LLM integration, prompt engineering, vector search Computer Vision: OCR systems, OpenCV preprocessing, image-to-data extraction ML Development: Regression, classification, clustering, model deployment Backend Engineering: FastAPI, Docker, REST APIs, scalable ML endpoints I don't chase Kaggle scores. I build systems that ship. MY STACK: Languages: Python, SQL GenAI/NLP: LangChain, Hugging Face, Transformers, Prompt Engineering ML/AI: Scikit-Learn, TensorFlow, PyTorch, Pandas, NumPy Computer Vision: OpenCV, Tesseract OCR Frameworks: FastAPI, Flask, Streamlit DevOps: Docker, Git/GitHub, CI/CD, REST APIs WHAT I'M LOOKING FOR: AI/ML Engineer, Data Scientist, or GenAI Engineer roles starting 2026. I want to join teams that: Ship ML/AI systems to production, not just POCs Value clean code and reproducibility Measure impact in business metrics, not just model accuracy If you're building AI that matters, let's talk. shindesarthak36@gmail.com

Location

Pune/Pimpri-Chinchwad Area

AI stack

antigravity

claude

codex

Work experience

AI/ML & Data Science Intern at Tata Motors, Oct 2025 - Jan 2026 · 4 mos, Pimpri Chinchwad, Designed and deployed a production-grade OCR and ML analytics pipeline that automated manual trainee assessment workflows — processing 1,000+ unstructured paper records and cutting evaluation cycle time by 70-80%. OCR PIPELINE DEVELOPMENT - Architected an end-to-end digitization system (OpenCV + Tesseract + Python), converting 1,000+ paper records into structured, analysis-ready datasets - Engineered a custom preprocessing chain (Gaussian denoising → adaptive binarization → multi-pass normalization), improving OCR accuracy by 25% - Eliminated the fully manual data entry process — the largest bottleneck in the evaluation workflow MACHINE LEARNING & ANALYTICS - Built and validated regression and K-Means clustering models in Scikit-Learn with 5-fold cross-validation - Performed EDA and feature engineering across 5+ training cohorts, improving training effectiveness scores by 15% - Applied statistical validation to ensure model reliability before handoff to reporting PRODUCTION DEPLOYMENT & ENGINEERING - Deployed a Dockerized KPI reporting platform (FastAPI + REST APIs), shifting dashboards for 10+ stakeholders from weekly Excel updates to real-time access — eliminating a 7-day reporting lag - Built a modular, Git-versioned codebase with independently testable components for reproducibility and future scaling - Automated the full workflow end-to-end: ingestion → preprocessing → modeling → reporting IMPACT: Replaced a manual, paper-based process with a Dockerized, real-time ML system — cutting cycle time 70-80%, improving accuracy 25%, and giving 10+ stakeholders real-time visibility. TECH: Python • OpenCV • Tesseract • Scikit-Learn • FastAPI • Docker • Pandas • NumPy • Git • REST APIs

Education

Bachelor of Technology, Computer Engineering, Nutan College of Engineering & Research

Public links

github: https://github.com/sarthak7-securtiy