Computer Vision & GenAI ML Engineer

Jenish Barvaliya

*******************@*****.*** +91-951******* Surat, Gujarat, India LinkedIn Profile GitHub Portfolio PROFESSIONAL SUMMARY

AI/ML Engineer with 8+ months of experience architecting real-time Computer Vision systems, multi-modal GenAI engines, and scalable ML infrastructure. Demonstrated expertise in zero-latency visual inspection (DINOv2, YOLO, mss screen capture), rubric-grounded LLM evaluation, and high-performance web scraping pipelines using Scrapling. Proven ability to deploy FastAPI microservices on cloud GPU clusters (Yotta GPU), orchestrate asynchronous task pipelines, and implement production MLOps practices.

TECHNICAL SKILLS

Computer Vision & Foundational Models:

DINOv2 & DINOv3, YOLO, OpenCV, Multi-Head Neural Networks, Feature Encoding, Real-Time Capture

GenAI & Multi-Modal NLP:

LLM Evaluation Engines, Rubric Scoring, Document OCR, Smart Layout Analysis, Prompt Engineering

Web Scraping & Data Pipelines:

Scrapling, Data Preprocessing, Content Classification MLOps, Cloud & Backend:

Yotta GPU Cloud, FastAPI Microservices, Model Serving & Deployment, Async Queues (Celery/Redis), Docker, Django REST, PyTorch, Python

PROFESSIONAL EXPERIENCE

AI/ML Engineer

Krira AI • Surat, Gujarat, India • kriraai.com

January 2026 – August 2026

Architected end-to-end real-time visual inspection pipeline (Positional Printing QC Automation); integrated high-speed screen capture, fabric localization via YOLO, DINOv2 & DINOv3 visual encoders, and a 5-head neural architecture emitting hardware signals (<0.5s latency), cutting fabric waste by 35%+ ( 2.9 Cr annual savings) Engineered asynchronous multi-modal evaluation engine (OSM AI Exam Grading) combining document OCR, intelligent CV page stitching/splitting, and rubric-grounded LLM handwritten answer scoring, collapsing grading turnaround from weeks to minutes per paper

Developed automated web scraping & SEO intelligence pipeline using Scrapling and NLP classification, automating backlink discovery, domain authority verification, and content categorization for growth teams Orchestrated batch model inference across Yotta Cloud GPU clusters via high-throughput FastAPI microservices and async task queues; actively expanding production MLOps deployment workflows and model serving pipelines AI/ML Engineer Intern

Elite Technocrats • Surat, Gujarat, India

May 2025 – June 2025

Developed and deployed AI-powered documentation automation tool using Django and React.js, reducing manual documentation time by 60%

Architected robust data pipeline for preprocessing and validating large-scale datasets, ensuring 99% data integrity for downstream NLP and ML workflows

TECHNICAL PROJECTS

SmartGriev AI-Powered Civic Complaint Management System github.com/jenish2917/smartgriev

Built civic grievance platform using Google Gemini AI supporting 12 Indian languages with voice input, image upload, and GPS location capture

Integrated YOLOv8 for real-time image defect analysis alongside multilingual NLP sentiment analysis (87% accuracy) Developed full-stack application with Django REST Framework and React TypeScript with 140+ Playwright E2E tests EasyArch AI-Powered Architectural Design Automation github.com/jenish2917/EasyArch

Built AI platform using DINOv2 Vision Transformer for floorplan vectorization into DWG format with 95%+ category accuracy Implemented PyTorch multi-task architecture with progressive resolution training (128px 384px), achieving a 70% loss improvement

EDUCATION

Bachelor of Technology (B.Tech) in Information Technology Sarvajanik College of Engineering and Technology • Surat, Gujarat, India Expected Graduation: May 2026

Cumulative GPA: 9.34/10.0

CERTIFICATIONS

Foundation Course on Green Skills and Artificial Intelligence Skills4Future Program • AICTE & Edunet Foundation

Completed: March 2025

View Certificate

GenAI Powered Data Analytics Job Simulation

Forage • Data Analytics & AI Collections Strategy

Completed: June 2025

View Certificate

Contact this candidate