LLM Evaluation Engineer & Full-Stack Developer

Jash Parmar

# ********@*****.*** +91-931******* ï LinkedIn § GitHub Portfolio Ð Codolio

Education

Bachelor of Technology, Computer Science Engineering 2022 – 2026 Indian Institute of Information Technology, Kota CGPA: 6.87/10 Experience

LLM Trainer, Handshake AI Jul 2026 – Present

Remote

• Designed and developed complex coding benchmark tasks for evaluating Large Language Model (LLM) agents in realistic terminal-based software engineering environments.

• Built end-to-end evaluation pipelines including Docker-based environments, automated test suites, reference implementations, and deterministic verification frameworks.

• Authored software engineering benchmark tasks spanning debugging, systems programming, data process- ing, and infrastructure workflows.

Software Development Engineer Intern, Xelron AI Feb 2026 – June 2026 Remote

• Contributed to AI evaluation projects including Terminus and Finch, improving LLM performance through prompt engineering, response analysis, and benchmark-driven evaluation.

• Worked on Project 1307, a collaboration with Uber AI Solutions, creating high-quality prompts and evaluating Gemini model responses to improve instruction following and reasoning capabilities.

• Collaborated with distributed teams to deliver high-quality AI training data and evaluation workflows for enterprise LLM applications.

Full-Stack Developer Intern, Flitzen Technologies UK Limited May 2025 – July 2025 Remote

• Led full-stack development of responsive e-commerce platform for daily hole-in-one golf competitions; en- abled £10 entries/subscriptions for unlimited access to £10k cash prizes across UK clubs.

• Designed intuitive UI with bold navigation (Competitions, Verified Play, Winners’ Circle, FAQs), contact forms, QR integration, and mobile layouts to boost engagement and conversions.

• Developed subscription features and Winners’ Circle gallery with verified win videos to enhance user reten- tion and build community.

Projects

PatternSense – AI-Powered DSA Pattern Detector APK Link GitHub React Native (Expo), TypeScript, Node.js, Express.js, Google Gemini API, Render

• Developed an AI-powered DSA pattern detection app using React Native and Gemini API, generating examples, approaches, and complexity analysis.

• Combined heuristic logic with LLM reasoning to support 20+ DSA patterns, deployed with a production- ready backend on Render.

LLM Powered Smart Screen Recorder GitHub

Python, Tesseract OCR, Sentence Transformers, ChromaDB, Google Gemini API, OpenCV

• Built Python RAG pipeline with screen capture, OCR, semantic embeddings, vector database storage, and Gemini-powered Q&A for querying visual work history. Technical Skills

• Languages: C++, Python, JavaScript, TypeScript, Java

• Web Technologies: ReactJS, React Native (Expo), Node.js, Django, HTML5, CSS3, Bootstrap, Tailwind

• Databases: MySQL, PostgreSQL

• AI/ML: Google Gemini API, Prompt Engineering, RAG, LLM Evaluation, AI Benchmarking, Sentence Transformers, Semantic Search

• Tools & Platforms: Git, GitHub, Docker, Linux, VS Code, Expo, Postman, Jupyter Notebook, Render, Vercel, Cursor AI

Contact this candidate