Developed a progressive web application integrating 3 major platforms (Codeforces, CodeChef, LeetCode) with automated cron jobs to fetch contest data, serving 500+ users with real-time updates. Deployed a smart notification system using Nodemailer to deliver scheduled email reminders, featuring secure authentication flows with OTP verification and cross-platform friend tracking.
"}},{"node":{"frontmatter":{"title":"Agentic Browser","cover":{"childImageSharp":{"gatsbyImageData":{"layout":"constrained","placeholder":{"fallback":"data:image/png;base64,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"},"images":{"fallback":{"src":"/static/90a62959dec7b1faca2ce5105570c4be/c13f1/cover.png","srcSet":"/static/90a62959dec7b1faca2ce5105570c4be/29e34/cover.png 175w,\n/static/90a62959dec7b1faca2ce5105570c4be/27866/cover.png 350w,\n/static/90a62959dec7b1faca2ce5105570c4be/c13f1/cover.png 700w,\n/static/90a62959dec7b1faca2ce5105570c4be/26347/cover.png 1400w","sizes":"(min-width: 700px) 700px, 100vw"},"sources":[{"srcSet":"/static/90a62959dec7b1faca2ce5105570c4be/40567/cover.avif 175w,\n/static/90a62959dec7b1faca2ce5105570c4be/5e660/cover.avif 350w,\n/static/90a62959dec7b1faca2ce5105570c4be/f5920/cover.avif 700w,\n/static/90a62959dec7b1faca2ce5105570c4be/35191/cover.avif 1400w","type":"image/avif","sizes":"(min-width: 700px) 700px, 100vw"},{"srcSet":"/static/90a62959dec7b1faca2ce5105570c4be/1dac5/cover.webp 175w,\n/static/90a62959dec7b1faca2ce5105570c4be/4d0bf/cover.webp 350w,\n/static/90a62959dec7b1faca2ce5105570c4be/a646e/cover.webp 700w,\n/static/90a62959dec7b1faca2ce5105570c4be/1f347/cover.webp 1400w","type":"image/webp","sizes":"(min-width: 700px) 700px, 100vw"}]},"width":700,"height":455}}},"tech":["Python","LangChain","React","TypeScript","WebSockets"],"github":"https://github.com/tashifkhan/agentic-browser","external":null},"html":"Architected a Model Agnostic Chrome extension using Deep Agent architecture to orchestrate complex browser automations, improving task decomposition success rates by 40% over traditional scripts. Implemented a Semantic DOM Compression pipeline via WebSockets to overcome LLM token limits, reducing API latency by 79% and minimizing memory overhead during complex DOM manipulation tasks.
"}},{"node":{"frontmatter":{"title":"Converse","cover":{"childImageSharp":{"gatsbyImageData":{"layout":"constrained","placeholder":{"fallback":"data:image/png;base64,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"},"images":{"fallback":{"src":"/static/35cba1fdd0a001074eea242c6e910bac/c13f1/cover.png","srcSet":"/static/35cba1fdd0a001074eea242c6e910bac/29e34/cover.png 175w,\n/static/35cba1fdd0a001074eea242c6e910bac/27866/cover.png 350w,\n/static/35cba1fdd0a001074eea242c6e910bac/c13f1/cover.png 700w,\n/static/35cba1fdd0a001074eea242c6e910bac/26347/cover.png 1400w","sizes":"(min-width: 700px) 700px, 100vw"},"sources":[{"srcSet":"/static/35cba1fdd0a001074eea242c6e910bac/40567/cover.avif 175w,\n/static/35cba1fdd0a001074eea242c6e910bac/5e660/cover.avif 350w,\n/static/35cba1fdd0a001074eea242c6e910bac/f5920/cover.avif 700w,\n/static/35cba1fdd0a001074eea242c6e910bac/35191/cover.avif 1400w","type":"image/avif","sizes":"(min-width: 700px) 700px, 100vw"},{"srcSet":"/static/35cba1fdd0a001074eea242c6e910bac/1dac5/cover.webp 175w,\n/static/35cba1fdd0a001074eea242c6e910bac/4d0bf/cover.webp 350w,\n/static/35cba1fdd0a001074eea242c6e910bac/a646e/cover.webp 700w,\n/static/35cba1fdd0a001074eea242c6e910bac/1f347/cover.webp 1400w","type":"image/webp","sizes":"(min-width: 700px) 700px, 100vw"}]},"width":700,"height":455}}},"tech":["React","Express.js","Socket.io","MongoDB"],"github":"https://github.com/vansh1293/Converse","external":"https://converse.vansharora.codes/"},"html":"Constructed an interactive chat interface employing React hooks and Socket.io, utilizing 11+ ShadCN UI elements to handle concurrent messaging. Engineered a real-time messaging API with secure file uploads for audio messages, implemented JWT-based authentication and optimized database storage for large binary assets, reducing message latency by 15%.
"}}]}}}