Ashwin Thakur

Computer Science undergraduate with hands-on experience in machine learning, computer vision, and NLP. Proven ability to design end-to-end AI/ML solutions — from data ingestion and model training to production-ready APIs. Skilled in Python, OpenCV, TensorFlow, Scikit-learn, and LLM-based retrieval systems. Seeking roles in ML engineering, AI development, or data science.

Software Engineering Intern at UCO Bank (2026-07 – 2026-09)

• Building a face recognition application that identifies customers in real time and automatically opens their corresponding profile with complete account details.
• Working on the facial detection and recognition pipeline to match live camera input against stored customer records for fast, touchless identification.
• Integrating the recognition system with backend customer data records to streamline retrieval of customer files at the point of service.

Project: Agentic SOP Generator at UCO Bank (2026-07 – 2026-09)

Built an offline, multi-agent tool that turns unstructured banking policies and regulatory circulars (PDF/DOCX/scanned) into structured, publication-ready SOPs — combining OCR ingestion, local LLM reasoning, and automated compliance checks.

• Implemented multi-threaded OCR ingestion (RapidOCR/Tesseract) and local LLM inference (Ollama, fine-tuned LoRA adapters) for zero external API dependency.
• Built an automated RACI matrix generator and compliance auditor, and a document engine that outputs formatted Word/Markdown SOPs with embedded flowcharts.

Generative AI & LLM Application Development Intern at Sparrow Softech Pvt. Ltd. (2026-06 – 2026-06)

• Completed a one-month internship developing AI-powered applications using Generative AI concepts and LLM-based workflows.
• Applied prompt engineering techniques and interaction design principles while building AI assistants and intelligent automation solutions.
• Used NLP techniques for text analysis and information extraction, and built Python-based AI applications integrating machine learning components.

Automation Intern at Tata Steel Ltd. (2025-05 – 2025-06)

• Engineered a deep learning computer vision system to automate industrial bar counting, eliminating operator errors and reducing manual inspection time.
• Developed OpenCV-based image processing pipelines in Python for real-time production-line monitoring, improving throughput tracking accuracy.

B.Tech in Computer Science Engineering – KIIT University (2027-05)