Achievements & Recognition
Award-winning performance in competitions and hackathons
Agentic RAG that maps developer profiles to IEEE assets for targeted skill-gap remediation. Ingests multi-source user data, analyzes skill gaps against live industry demands, and retrieves tailored learning paths, standards, and mentor matches from IEEE's ecosystem.
3rd Place Overall – Pearl Hacks 2026 (MLH, UNC Chapel Hill) | Best Use of Solana. Built an AI-powered Health + Fintech platform that brings proactive clarity to healthcare costs — before visits, after claims, and across financial planning — while enabling community-backed support through blockchain.
▶ Watch Careless Demo
Built an AI Agent for holistic student performance analysis, integrating multiple data sources to provide comprehensive insights for educators and students.
Developed a semantic NASA Knowledgebase Engine for publication exploration, enabling researchers to efficiently navigate and discover relevant scientific literature.
Conducted by IBM, demonstrating excellence in data-driven problem-solving and analytics
National-level hackathon conducted by Government of India, competing among thousands of teams nationwide
Certifications
IBM Developer Skill Network - Comprehensive training in data analysis, machine learning, and statistical modeling
Advanced certification covering Hadoop, Spark, and large-scale data processing techniques
End-to-end fullstack development with Java, Spring Boot, and modern frontend frameworks
Scrum methodology, sprint planning, and collaborative development practices
Foundational cloud concepts, service models, and deployment strategies
CISCO certification covering network security, threat detection, and risk management
Professional Training
AWS fundamentals, cloud architecture, and service deployment best practices
Django, Flask, and modern Python web frameworks for scalable applications
Building intelligent agents, LLM integration, and conversational AI systems
Advanced problem-solving techniques, algorithm optimization, and computational complexity analysis
Research Ongoing
Towards research paper
- Building an end-to-end ML pipeline for EEG (32-channel) and fNIRS data, benchmarking 8+ imputation algorithms (matrix completion, tensor decomposition) to address missing neural signals.
- Engineering temporal, spectral, and spatial features from multimodal neural data for cognitive state classification; applying statistical analysis and data visualization to evaluate signal quality.
- Implementing a Conformer-based deep learning model (PyTorch) and performing hypothesis testing and statistical evaluation to compare signal reconstruction and classification performance across modalities.