Computer Science undergraduate with a strong foundation in software engineering, edge AI, and distributed systems. I enjoy designing, implementing, and optimizing end-to-end products, from fine-tuning large vision-language models to building decentralized systems and high-performance backends. I like ambitious ideas, clean execution, and solving hard technical problems under pressure.

• University of Lucknow — B.Tech in Computer Science and Engineering (Artificial Intelligence)
• Expected Graduation: May 2027
• Current CGPA: 8.5 / 10.0

Data Analytics and Machine Learning Intern | Jun 2025 - Aug 2025

• Designed and implemented an Edge AI multimodal computer vision system using architectures such as EfficientDet, YOLO, and DETR, enabling alert generation in under 300 milliseconds.
• Built geospatial mapping and region-based decision logic for spatially aware analysis and autonomous zone-level actions.
• Optimized multimodal feature fusion pipelines combining vision and sensor signals, improving detection accuracy by 11.9% over baseline models.

Python, IPFS, Web3, Distributed ML

• Led team Star Busters to secure 1st Place in the Web3 track at RIFT 2026, outperforming 8,000+ participants across 4 cities and winning a $250 Algorand grant.
• Engineered a decentralized operating system using IPFS and quantum-encryption-inspired architecture to crowdsource idle GPU and CPU compute securely.
• Reimagined resource-heavy cryptographic workflows by shifting the system toward decentralized ML training infrastructure for better compute utilization.

Python, LLaMA Vision, API Integration

• Built an assistive wearable system powered by a fine-tuned LLaMA Vision Language Model trained on egocentric data for high-accuracy environmental understanding.
• Developed an agentic navigation module using real-time sensor fusion and routing APIs for autonomous guidance and obstacle avoidance.
• Reduced edge inference latency by 25%, improving real-time audio feedback and hazard detection.

Python, GAT, Cython, FastAPI

• Engineered a real-time solar event detector combining AI and physics-based methods to analyze Coronal Mass Ejection behavior.
• Accelerated preprocessing by 15x with Cython optimizations on large space observation datasets.
• Deployed a hybrid GAT + Transformer-BiLSTM + Isolation Forest pipeline for robust high-speed anomaly detection.

• CME Particle Analysis: Research selected under an ISRO initiative, focused on AI-driven analysis of solar activity relevant to Aditya-L1 payload measurements.
• Adaptive Hybrid DDoS Mitigation System: Accepted at ICICT 2026, London, UK, featuring entropy-based anomaly detection and client-puzzle defense mechanisms.
• Assistive IoT Navigation: Published work on end-to-end hardware-software architecture for low-latency navigation support for visually impaired users.
• Competitive Problem Solving: Reached a peak LeetCode rating of 1737.
• Hackathons: 4x winner, including RIFT 2026 Web3 Track and Techathon 2.0.

• Strong interest in edge AI, distributed systems, decentralized infrastructure, and applied research
• Experience spanning graphics design, animation, video editing, 3D workflows, and technical storytelling
• Comfortable leading teams, shipping under pressure, and translating complex ideas into polished outcomes