🎓 MS in Computer Software Engineering @ Northeastern University (GPA: 4.0/4.0)

I am a graduate student at Northeastern University, pursuing a Master's in Computer Software Engineering, where I am deepening my expertise in software engineering, cloud architecture, and intelligent systems

A dynamic, full-stack Software Engineer with expertise in building scalable applications, AI-integrated systems, and cloud-native architectures. Passionate about delivering software that is not just functional — but maintainable, testable, and built to scale. My experience spans full-stack development, AI-powered application engineering, and end-to-end software delivery.

💡 What sets me apart: I understand the entire software development lifecycle — designing, building, and shipping features that work flawlessly at scale. From backend systems to AI-powered features to polished frontends, I build clean, production-ready software end-to-end.

📍 Location: Seattle, WA 🎯 Seeking: Fall 2026 Software Engineering | AI Engineering | Full-Stack Development Internship opportunities

Master of Science in Computer Software Engineering - Information Systems Northeastern University, Seattle | 2025 - 2027 | GPA: 4.0/4.0

Relevant Coursework: Object-Oriented Design, Data Structures & Algorithms, Web Development, Software Engineering

Bachelor of Engineering (Honors) - Computer Science Rajiv Gandhi Proudyogiki Vishwavidyalaya, India

Tech Stack: FastAPI · Python 3.11 · Groq (LLaMA 3.3-70B) · React · Vite · EDI 278

🔗 GitHub Repo

🏆 1st Place Winner — AI Agents Hackathon, hosted by OSS4AI in collaboration with Beta Fund and Gravitational Ventures. Built solo in 36 hours.

An AI agent that automates healthcare prior authorization end-to-end — scanning payer policies, analyzing medical records, submitting via EDI 278, and auto-drafting appeals on denial. Built to attack the $19.7B annual administrative burden U.S. healthcare carries for a process that still runs on fax machines.

Key Highlights:

• 🤖 Real agentic tool-use loop — the LLM decides which tool to call next, observes the result, and continues until the workflow completes (no hardcoded sequences)
• 🔄 6-tool end-to-end workflow — policy scanning → medical record analysis → form filling (96% completeness) → EDI 278 submission → status tracking → auto-appeal drafting
• 📝 Denial auto-appeal — reads the denial code, pulls supporting clinical evidence, and generates a CMS-compliant appeal letter in seconds (a 2–3 day manual process compressed to real time)
• 🏥 HIPAA-aware, CMS 2026 compliant across 5 payers — Aetna, BCBS, UHC, Cigna, Humana — covering ~60% of the U.S. insured population
• ⚡ Built and shipped solo in 36 hours, from problem framing to a working demo

💬 "Tackles a massive, well-defined problem with genuine agentic design... the denial → auto-appeal flow demonstrates product thinking beyond the happy path." — Hackathon Judge

React Frontend (Vite)
        ↓ HTTP + polling
FastAPI Backend
        ↓
Agentic Runner (Groq · LLaMA 3.3-70B)
   LLM → tool_use → observe → repeat
        ↓
Healthcare Vertical (6 tools)
   EDI 278 · CMS 2026 · HIPAA-aware

Tech Stack: Python · FastAPI · React · Claude API · RAG · FAISS · Streamlit · Docker

🔗 GitHub Repo

An agentic AI music recommendation system that understands natural language — tell it your mood, context, and vibe, and it curates a personalized playlist with AI-powered reasoning.

Key Highlights:

• 🤖 Claude API parses natural language into structured preferences via Pydantic schema
• 🎯 12-dimensional weighted scoring engine across mood, energy, and genre dimensions
• 🔄 Multi-step agentic workflow — intent parsing → retrieval → scoring → validation → explanation
• 📊 87% recommendation confidence with mood/context alignment validation
• 🎧 5 specialized profiles — study, party, workout, relax, night
• ✅ Reliability harness with synthetic test cases — 3/4 alignment cases passed
• ⚡ Prompt caching to reduce API latency and token cost

Natural Language Request
        ↓
Claude Intent Parser (Pydantic Schema)
        ↓
RAG Candidate Retriever
        ↓
12-Dimensional Scoring Engine
        ↓
Diversity Filter + Validation
        ↓
Confidence Score + Claude Explanation
        ↓
Curated Playlist Output

Tech Stack: Go · Next.js · TypeScript · PostgreSQL · Redis · WebSockets · Kubernetes · AWS (EKS, RDS, ElastiCache, S3, CloudFront) · Claude API · Docker · Tailwind CSS

🔗 GitHub Repo

A production-grade, full-stack social media platform built from the ground up — backend, frontend, infrastructure, and AI features all implemented independently.

Key Highlights:

• ⚡ 40% latency reduction via Redis + AWS ElastiCache caching layer
• 🔌 1,000+ concurrent WebSocket connections for real-time notifications (likes, follows, replies, mentions)
• 🤖 Claude API powers a hybrid feed ranking algorithm blending content quality, recency decay, and engagement signals
• 🗄️ PostgreSQL with GIN full-text indexes and denormalized counts for read performance
• ☁️ AWS EKS with Kubernetes HPA, RDS, ElastiCache, S3 + CloudFront for media delivery
• 🔐 JWT authentication + bcrypt hashing for secure multi-user access

Go REST API → PostgreSQL + Redis → WebSocket Hub
     ↓               ↓                  ↓
JWT Auth      ElastiCache Cache    1000+ clients
     ↓               ↓                  ↓
Claude API    Feed Ranking Algo    Real-time alerts
     ↓
AWS EKS + Kubernetes HPA

Tech Stack: React · Node.js · PostgreSQL · Redis · WebSocket · Claude API · Docker · React Native · WebXR · GitHub Actions

🔗 GitHub Repo

A full-stack AI plant care platform with multimodal plant identification, real-time care alerts, and AR visualization.

Key Highlights:

• 🔐 Secure multi-user REST API with JWT authentication, bcrypt hashing, PostgreSQL + Redis caching
• 🤖 Claude API across 4 AI endpoints — plant identification, health diagnosis with confidence scoring (0–100%), persistent chatbot memory, and real-time WebSocket push alerts
• 📱 Cross-platform React Native + Expo mobile app with camera-based diagnosis
• 🌐 WebXR/AR visualization using ARCore/ARKit for immersive plant viewing
• 🐳 Dockerized full stack with GitHub Actions CI/CD pipeline and integration test coverage
• 🌱 Plant catalog management across 12+ species with watering schedule automation

Tech Stack: Python · FastAPI · Amazon Bedrock · Apify · Box

🔗 GitHub Repo

Built at the CascadiaJS Hackathon as part of a 3-person team — an agentic assistant that hunts down the best deals across the web and turns scattered listings into a single, ranked recommendation.

Key Highlights:

• 🤖 Amazon Bedrock-powered reasoning — the agent plans searches, evaluates listings, and ranks deals against user intent
• 🕸️ Apify-driven web scraping pipeline for live product and pricing data across multiple retail sources
• 📦 Box integration for storing and organizing extracted deal data and generated reports
• ⚡ FastAPI backend orchestrating the scrape → analyze → rank → recommend workflow
• 👥 Built collaboratively as a team of 3 under hackathon time constraints

User Query
        ↓
FastAPI Backend
        ↓
Apify Scrapers → Live Listings
        ↓
Amazon Bedrock Agent (plan → evaluate → rank)
        ↓
Box (storage) → Ranked Deal Recommendations

Tech Stack: Spring Boot · React · MongoDB · AWS S3 · Kafka · Claude API · FAISS · Apache Tika · Kubernetes · Docker

🔗 GitHub Repo

ctrl+F finds words. DocuMind finds meaning.

An AI-powered document intelligence platform that lets you query your entire document library using natural language — powered by RAG, FAISS vector search, and Claude AI.

Key Highlights:

• 🔍 85% NLP semantic similarity match via RAG + FAISS vector indexing
• 🤖 Claude API for keyword extraction, summarization & document Q&A
• 📄 Multi-format support — PDF, DOCX, images via Apache Tika
• ⚡ Real-time Kafka pipeline for event-driven document ingestion
• 🔒 AWS S3 with AES-256 encryption + JWT authentication
• 🚀 Kubernetes deployment with Docker + GitHub Actions CI/CD

Upload → Tika extracts text
       → Kafka processes pipeline
       → Claude extracts keywords
       → FAISS indexes vectors
       → MongoDB stores metadata
       ↓
Query  → FAISS similarity search
       → Claude generates answer
       → React displays results

HCL America (Client: BECU) | Seattle, WA | Jul 2024 – Apr 2025

• 🏗️ Developed scalable automation infrastructure using Selenium WebDriver (Java) for 2 enterprise applications in a cloud-native microservices architecture
• ⚙️ Established cloud instances using Terraform (IaC) and set up Jenkins CI/CD pipelines to automate deployment workflows across distributed engineering teams
• ✅ Validated REST APIs with Postman and executed SQL database testing across 100+ business requirements

Sapiens | Bangalore, India | May 2021 – Dec 2022

• 🔧 Built a Java automation framework with TestNG and Page Object Model embedded into a CI/CD pipeline, achieving 80% regression coverage
• 🤝 Partnered with cross-functional Agile teams to surface edge cases, resolve defects via Jira, and coached 2 junior engineers in SDLC best practices

Deloitte | Hyderabad, India |

• 🏗️ Built a scalable automation framework from scratch using Java/Selenium with behavior-driven development (BDD); collaborated on code reviews and created internal productivity tools across 3 projects
• 📝 Authored and maintained technical documentation; applied object-oriented design principles to identify edge cases, debug issues, and deliver reliable software across Agile sprints

💼 Open to: Software Engineer | AI Engineer | Full-Stack Developer Internships (Fall 2026)

⭐ If you like my work, consider giving a ⭐ to my repositories!

🚀 Let's build something amazing together!