Devin Gundaniya

I’m a Software Engineer and AI Evaluation Engineer with 4+ years of experience building scalable web applications, SaaS platforms, and AI-powered solutions. I combine modern full-stack engineering with a quality-first mindset, and I recently expanded into AI Evaluation Engineering by designing and validating benchmarks for frontier LLMs.

At Handshake AI (Remote · San Francisco, California, United States), I built realistic software engineering benchmarks to evaluate frontier AI coding agents. I created terminal-based debugging and repository repair tasks, developed deterministic verification criteria, and automated evaluation workflows. I also produced reproducible Linux/Docker environments and analyzed AI-generated solutions to identify reasoning failures, implementation errors, and edge cases.

Before that, at Desilo Technologies Privet. Limited, I started as a Junior Software Developer and progressed into a Full Stack Developer. I focused on React.js and Redux to build responsive, user-centric interfaces, then scaled to full-stack development with React.js, Node.js, Express.js, MongoDB, and Python—while integrating REST APIs and maintaining high-performance MERN-style applications.

Across my projects, I keep pushing toward practical Generative AI and strong developer tooling—whether that’s AI-powered customer success dashboards, real-time analytics and automation, or an AI CRM using the OpenAI API and a Socket.io chatbot. I care deeply about software quality, reproducibility, and building systems that are reliable under real debugging and edge-case pressure.