Greater Bengaluru Area
Most CS students build projects. I build and ship them — deployed, documented, and open-source. I'm an AI Engineer and full‑stack developer who cares about real users, real traffic, and real constraints. Over the past 2+ years, I've built systems that range from Android apps running on-device ML to backend services handling hundreds of concurrent requests. Some highlights that separate me from typical "course + toy project" profiles: - Built a multi-threaded C++ Deep Packet Inspection (DPI) engine with load balancing and fast-path architecture, parsing real network traffic and classifying applications at scale. - Shipped a production-ready digital watermarking web app (Flask + Docker) using DWT, SVD, and meta-heuristic optimization, complete with live demo, metrics (PSNR, SSIM, MSE, NCC), and user-friendly UI. - Designed and deployed REST APIs (Node.js, Express, MongoDB/PostgreSQL) that serve 200+ concurrent requests with sub-100ms response times using indexing, caching, and robust error handling. On the ML side, I combine strong fundamentals with hands-on work: - Selected for Amazon ML Summer School 2025 (top ~5%), with 80+ hours of supervised/unsupervised learning, deep neural networks, LLMs, reinforcement learning, and causal inference. - Implemented end-to-end ML workflows: data prep, model training, evaluation, and serving, including experiments with generative AI and LLM-based tooling. Tech stack: Python, PyTorch, TensorFlow, Flask, Node.js, Express, Docker, AWS, Firebase, Kotlin/Android, PostgreSQL, MongoDB, CI/CD, REST APIs, multithreading, networking. I'm currently looking for AI Engineer / ML Engineer / Software Engineer roles where I can own features end-to-end: from idea, to model, to API, to production deployment. If you're building systems at the intersection of AI and high-performance backend engineering, I'd love to connect.
• Engaged in digital inspection, asset management, and workflow automation at Spectent™. • Developed scalable backend systems to tackle operational challenges effectively. • Collaborated with cross-functional teams to enhance product offerings and streamline processes. • Gained hands-on experience with innovative technologies in a dynamic startup environment.
At Ethara AI, I focused on enhancing the performance of Large Language Models through post-training workflows. My role involved evaluating model responses to detect inconsistencies and hallucinations, which was crucial for improving output quality. I designed various prompt strategies and validation techniques to ensure the reliability of the models, contributing to a more accurate response generation process.
• Evaluated AI-generated outputs using structured analytical workflows. • Identified logical inconsistencies, formatting issues, and edge cases in AI responses. • Worked in fast-paced environments requiring analytical thinking, workflow accuracy, and attention to detail. • Collaborated within process-driven systems for response validation and quality evaluation.
• Selected among the top 5% of applicants nationwide for Amazon's most competitive ML training program, outperforming thousands of CS students across India • Completed 80+ hours of intensive training directly led by Amazon Scientists, covering LLMs, deep neural networks, reinforcement learning, causal inference, and sequential learning for time-series prediction • Scored in the top tier of program assessments across 8 modules, demonstrating applied understanding of how ML systems operate at Amazon's scale in e-commerce and cloud infrastructure • Built foundational expertise in generative AI and model fine-tuning techniques, with direct exposure to real-world deployment considerations from practicing Amazon researchers • Engaged in live technical Q&As with senior Amazon ML Scientists, bridging theory with production-grade applications in recommendation systems and demand forecasting
• Architected and deployed 5+ production-ready REST APIs using Node.js, Express.js, and MongoDB, each with comprehensive input validation, error handling, and Swagger/OpenAPI documentation • Cut API regression bugs by 40% through systematic test automation using Keploy CLI — generating, replaying, and maintaining test suites across the full development lifecycle • Built RESTful services handling 200+ concurrent requests with sub-100ms response times, achieved through efficient database indexing and query optimization strategies • Delivered all 12 assignments ahead of schedule with zero revision requests, maintaining the highest completion rate in a cohort of 20+ fellows across India • Contributed 3 merged pull requests to Keploy's open-source codebase, improving CLI documentation clarity and increasing test case reliability for downstream users • Led weekly peer code reviews and pair programming sessions, identifying recurring API design anti-patterns and sharing fixes across the cohort