AI/ML Engineer & Full-Stack Developer specializing in agentic RAG, edge ML deployment, and production-ready AI systems.
About
I'm a Computer Science undergraduate (AI/ML) at UPES Dehradun with a 9.48 CGPA, focused on building systems that translate ML models into reliable, production-ready software.
I prioritize practical engineering: clear architecture, measurable performance, and maintainable code. Recent work spans cloud-first AI workflows, edge inference on constrained hardware, and data-driven product development.
Business Analyst Intern
Kailapira · Jun 2025 – Jul 2025
Technical Skills
Languages
AI / ML
Frameworks
Infrastructure
Design
Selected Work
Each project follows a clear path: identify a problem, engineer a solution and measure the impact.
Career guidance lacks personalization at scale
Built an agentic RAG platform with Amazon Bedrock that delivers context-aware career recommendations through multi-step retrieval and response orchestration.
Impact: End-to-end AI pipeline from document retrieval to personalized guidance delivery
Manual expense entry is tedious and error-prone
Developed an OCR-driven mobile app that automatically extracts and categorizes receipt data using TesseractOCR.
Impact: Reduced manual data entry effort by ~90%
No easy way to track how far you've watched through a YouTube playlist
Built a Firefox browser extension that overlays a persistent checklist on any YouTube playlist, letting users tick off videos as they watch and instantly see their progress.
Impact: Published on Firefox Add-ons (AMO) - turns any playlist into an interactive watch-tracker with zero configuration
Air quality monitoring needs low-latency edge inference
Built an edge-first monitoring pipeline using Jetson Nano with an ONNX-deployed LSTM model for real-time air quality predictions.
Impact: Sub-second inference latency on constrained edge hardware
Medical imaging classification needs transparent, reliable models
Combined adaptive median filtering with a tuned logistic regression model for chest X-ray classification with interpretable outputs.
Impact: 87.3% classification accuracy with full model transparency
Data structures like Binary Search Trees can be abstract and difficult to visualize without interactive tools
Built an interactive React-based visualizer with SVG rendering and step-based animations for tree operations like insert, search, and traversals.
Impact: Makes understanding complex tree algorithms intuitive through visual, step-by-step animated workflows
Recognition
State Winner - Udbhav Hackathon, Raj Bhavan Uttarakhand
View on LinkedInProject Expo Winner - State Foundation Silver Jubilee
View on LinkedInJoint Secretary, CSI UPES - Led workshops & hackathons for 200+ students
View on LinkedInPresented at 3rd International Young Researchers' Conclave (IYRC) 2025
View on LinkedInGet in Touch
Open to AI/ML engineering, full-stack, and systems internship opportunities. Drop me a line or connect on social.
Say Hello