- 🌟 Google Student Ambassador: Leading developer community growth, technical workshops, and flagship hackathons.
- 🔬 AI/ML Researcher & Developer: Specialized in low-latency LLM inference, edge AI, and production-grade Generative AI pipelines.
- 💻 Open Source Contributor: Actively contributing to real-world repositories in GSSoC '26.
- 🛡️ Cybersecurity Intern: Completed training focused on penetration testing, cryptography, and steganography under AICTE & IBM SkillsBuild.
- The Problem: Reviewer bias, scoring inconsistencies, and manual coordination bottlenecks in large-scale hackathons.
- The Solution: Built a multi-role platform using Next.js 14 (TypeScript & Tailwind CSS), a Node.js/Express API gateway, MongoDB Atlas, and a Python/FastAPI AI microservice powered by Sentence Transformers and Scikit-learn.
- Key Impact: Automated conflict-free judge allocation, balanced team clustering, duplicate submission detection, and real-time reviewer bias/anomaly flagging across the end-to-end event lifecycle.
- The Problem: Data exfiltration risks in highly sensitive, isolated network environments.
- The Solution: Engineered a zero-trust software-defined bridge using FastAPI for controlled data transit, React & Tailwind CSS for the monitoring interface, and Firebase Realtime Database for live state management.
- Key Impact: Maintained 99.9% isolation integrity and designed a modular API schema following OpenAPI standards for extensibility.
- The Problem: Lack of accessible employment infrastructure in low-connectivity rural areas.
- The Solution: Developed a privacy-first skill-matching engine using TensorFlow Lite for on-device edge inference, paired with FastAPI, Firebase, and a Dialogflow multilingual voice assistant.
- Key Impact: Reduced matching latency by 40% and delivered a scalable, voice-first architecture built for rural accessibility.
- The Problem: Insecure transmission of sensitive data over public channels.
- The Solution: Designed a pixel-level encryption and encoding tool using Python and OpenCV to conceal encrypted payloads inside carrier images.
- Key Impact: Achieved 100% lossless data recovery with zero detectable visual variance.
