A short story about what I work on and what I care about.
I'm a builder at heart. I gravitate toward problems that sit at the intersection of AI systems, backend infrastructure, and product engineering. The deeper the stack, the more interested I am.
My recent work spans LLM-powered products built on Gemini and modern TypeScript stacks, distributed browser isolation platforms with Go and Kubernetes, and medical imaging pipelines for neuroscience research achieving 92% accuracy. I care about reliability, observability, and code that holds up in production.
Outside of coursework at CSUF where I study Distributed Systems, Neural Networks, and Software Architecture. I ship side projects, dig into systems design, and stay current with the fast-moving AI tooling landscape.
Graduate work focused on systems, AI, and software engineering.
California State University, Fullerton·Fullerton, CA
Concentration: Intelligent Application Systems Engineering
Nirma University·Ahmedabad, India
A mix of distributed systems, AI-powered products, Web3, and research tooling. Each shipped end-to-end.
Graph-based Retrieval-Augmented Generation pipeline using a knowledge graph store instead of vector databases. Models entities and relationships as graph nodes and edges, enabling multi-hop traversal and structured context retrieval for grounded, relationship-aware LLM responses.
Distributed remote-browser platform — Docker-isolated Chromium sessions, Go orchestrator with CDP input forwarding, WebRTC streaming with WebSocket fallback, Kubernetes HPA autoscaling, coturn STUN/TURN relay, Prometheus + Grafana telemetry, and Playwright E2E coverage.
Full-stack GenAI interview coach with role-based questions, real-time feedback, and ratings via Gemini 2.5 Flash. Features live transcription, waveform visualization, word count, and timer using the Web Speech API. Protected user history and score analytics on Neon Postgres.
Cloud-native video platform with FFmpeg HLS pipeline, S3 adaptive streaming (360p / 720p / 1080p), and a pg-boss PostgreSQL job queue. Secured 8+ routes with Clerk JWT and RBAC. Powers AI-driven transcript search via Groq Whisper with tsvector full-text indexing and React heatmap analytics.
Blockchain-based identity and resume verification platform on Ethereum — Solidity smart contracts handle multi-role workflows between Applicants, Employers, and Institutions. Resumes stored on IPFS via Pinata with MetaMask wallet auth and dual-ledger sync across the blockchain and MongoDB.
Industry experience building web platforms and mobile apps shipping production code from day one.
Web Development Intern
Gujarat, India
Android Development Intern
Gujarat, India
Peer-reviewed ML research on medical imaging and deep learning for early disease detection.
Built a 3D CNN model for early Parkinson's Disease detection trained on PPMI clinical MRI scans with HPC multi-GPU infrastructure. Engineered a full preprocessing pipeline covering N4 bias correction, skull stripping, image registration, and normalization.
Reviewed 2D/3D MRI deep learning approaches for early PD detection across 50+ papers, covering preprocessing pipelines, model architecture trends, performance benchmarks, and open research gaps in the field.
A working toolkit, organized by where it fits in the stack.
Or connect with me on social media
