An AI-assisted incident-resolution prototype for cloud infrastructure. SmartOps brings telemetry, logs, diagnosis, suggested remediation, and human SRE review into one clear operating loop. Rahul led the AI/backend work, model and pipeline development, training, deployment, and proof-of-concept flow. The work reached the LTTS TECHgium 2026 national finalist stage as one of 34 teams from 62,000+ registrations.
"}},{"node":{"frontmatter":{"title":"Amazon ML Challenge 2025","cover":{"childImageSharp":{"gatsbyImageData":{"layout":"constrained","placeholder":{"fallback":"data:image/jpeg;base64,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"},"images":{"fallback":{"src":"/static/372e39e1e642a380ef26fcba8539176d/4155f/cover.jpg","srcSet":"/static/372e39e1e642a380ef26fcba8539176d/c3daa/cover.jpg 175w,\n/static/372e39e1e642a380ef26fcba8539176d/16cd1/cover.jpg 350w,\n/static/372e39e1e642a380ef26fcba8539176d/4155f/cover.jpg 700w","sizes":"(min-width: 700px) 700px, 100vw"},"sources":[{"srcSet":"/static/372e39e1e642a380ef26fcba8539176d/e2b90/cover.avif 175w,\n/static/372e39e1e642a380ef26fcba8539176d/881d8/cover.avif 350w,\n/static/372e39e1e642a380ef26fcba8539176d/ea1fc/cover.avif 700w","type":"image/avif","sizes":"(min-width: 700px) 700px, 100vw"},{"srcSet":"/static/372e39e1e642a380ef26fcba8539176d/eda49/cover.webp 175w,\n/static/372e39e1e642a380ef26fcba8539176d/c4210/cover.webp 350w,\n/static/372e39e1e642a380ef26fcba8539176d/cd33f/cover.webp 700w","type":"image/webp","sizes":"(min-width: 700px) 700px, 100vw"}]},"width":700,"height":316}}},"tech":["DeBERTa-v3-base","NLP regression","Hybrid pooling","Log-MAE","Validation"],"github":null,"external":"/archive/","cta":null},"html":"With Cyber Titans, Rahul led the model, training, submission, deployment, and AI/backend work for a price-prediction challenge. The final submission reached Rank 83 with 45.35 SMAPE, placing in the Top 2%. The core value was not a single model trick: it was a disciplined loop of diagnosis, controlled changes, and honest leaderboard validation.
"}},{"node":{"frontmatter":{"title":"Smart Glass Assistant","cover":{"childImageSharp":{"gatsbyImageData":{"layout":"constrained","placeholder":{"fallback":"data:image/jpeg;base64,/9j/2wBDABALDA4MChAODQ4SERATGCgaGBYWGDEjJR0oOjM9PDkzODdASFxOQERXRTc4UG1RV19iZ2hnPk1xeXBkeFxlZ2P/2wBDARESEhgVGC8aGi9jQjhCY2NjY2NjY2NjY2NjY2NjY2NjY2NjY2NjY2NjY2NjY2NjY2NjY2NjY2NjY2NjY2NjY2P/wgARCAAPABQDASIAAhEBAxEB/8QAGQAAAgMBAAAAAAAAAAAAAAAAAAUCAwQG/8QAFQEBAQAAAAAAAAAAAAAAAAAAAQD/2gAMAwEAAhADEAAAAcs+gRFYLgf/xAAcEAACAgIDAAAAAAAAAAAAAAAAAgERAxMEEiH/2gAIAQEAAQUCVX7bfbsZYY5WLWQf/8QAFxEBAAMAAAAAAAAAAAAAAAAAAAECEv/aAAgBAwEBPwGLNv/EABcRAQEBAQAAAAAAAAAAAAAAAAEAAhP/2gAIAQIBAT8Bcjc7/8QAGRAAAgMBAAAAAAAAAAAAAAAAAAEQITFB/9oACAEBAAY/AsLRUJLsf//EABwQAQADAAIDAAAAAAAAAAAAAAEAESExQVFhof/aAAgBAQABPyHe6fJnFS8kKhWMEc5KfcDyW4kKn//aAAwDAQACAAMAAAAQKA//xAAWEQEBAQAAAAAAAAAAAAAAAAAAAXH/2gAIAQMBAT8QpGH/xAAVEQEBAAAAAAAAAAAAAAAAAAAAcf/aAAgBAgEBPxAFv//EABsQAQEAAgMBAAAAAAAAAAAAAAERACExQYFR/9oACAEBAAE/EFyYK1OF5TBs2RXdeeYV0KiZSAjdEOEbwgrZDBPxrWf/2Q=="},"images":{"fallback":{"src":"/static/69f8f630ada4357e87a0e45fe153c012/10a6f/cover.jpg","srcSet":"/static/69f8f630ada4357e87a0e45fe153c012/37d55/cover.jpg 175w,\n/static/69f8f630ada4357e87a0e45fe153c012/67094/cover.jpg 350w,\n/static/69f8f630ada4357e87a0e45fe153c012/10a6f/cover.jpg 700w,\n/static/69f8f630ada4357e87a0e45fe153c012/9d298/cover.jpg 1400w","sizes":"(min-width: 700px) 700px, 100vw"},"sources":[{"srcSet":"/static/69f8f630ada4357e87a0e45fe153c012/ea096/cover.avif 175w,\n/static/69f8f630ada4357e87a0e45fe153c012/758bc/cover.avif 350w,\n/static/69f8f630ada4357e87a0e45fe153c012/23e81/cover.avif 700w,\n/static/69f8f630ada4357e87a0e45fe153c012/b929c/cover.avif 1400w","type":"image/avif","sizes":"(min-width: 700px) 700px, 100vw"},{"srcSet":"/static/69f8f630ada4357e87a0e45fe153c012/93fe5/cover.webp 175w,\n/static/69f8f630ada4357e87a0e45fe153c012/a9c60/cover.webp 350w,\n/static/69f8f630ada4357e87a0e45fe153c012/e1299/cover.webp 700w,\n/static/69f8f630ada4357e87a0e45fe153c012/64ed2/cover.webp 1400w","type":"image/webp","sizes":"(min-width: 700px) 700px, 100vw"}]},"width":700,"height":525}}},"tech":["Flutter","ESP32-CAM","Object detection","OCR","Edge-oriented design"],"github":null,"external":"/archive/","cta":null},"html":"An accessibility prototype that combines wearable visual input with a mobile experience for object, text, and face assistance. Rahul built the application and contributed to the system integration work. It is presented as an engineering prototype — not a claim that a demo replaces a production assistive device.
"}}]}}}