Hi, I'm
Arnav Khare
Software Engineer / AI Systems Builder
Informatics student at the University of Washington (GPA 3.71, graduating March 2026) specializing in full-stack development, AI/ML systems, and cloud infrastructure. Building Gnome, an IoT + AI plant care startup that won $20K at the Harriet Stephenson Business Plan Competition.
Scroll02. experience
Where I've Worked
Research Assistant
University of Washington
School of Built Environments · Seattle, WA
- ▹Uplifted the yearly activity report (YAR) system into a complete product applying SDLC principles — requirements gathering, design, development, testing, automation, and deployment — enabling streamlined faculty performance tracking
- ▹Implemented CRUD operations with MongoDB, managing faculty records and evaluation forms with 100% data integrity
- ▹Deployed application using Netlify and Windows VM
Software Engineering Intern
T-Mobile
Subscriber Event Listener Team · Bellevue, WA
- ▹Built an end-to-end AI-powered data visualization tool across 4 microservices leveraging Azure OpenAI, reducing graph query time from 1–2 hours to 5–10 seconds
- ▹Implemented FAISS vectorstore with LangChain for semantic search across 10M+ customer records, achieving sub-second query times; developed autonomous visualization agent generating 10+ graph types from CSV, Excel, and JSON
- ▹Designed Springboot end-to-end authentication system with bearer token security for subscriber lookup, collaborating with cross-functional teams
Full-Stack Software Engineering Intern
GOLS
360 Peer-Review Team · Seattle, WA
- ▹Resolved inefficient peer-feedback analysis by building a full-stack web app leveraging OpenAI API and MySQL to display key findings and automate comment management, reducing feedback processing time
- ▹Developed back-end API using Node.js to update the database with user information
- ▹Implemented 6 UI design sketches using HTML/CSS for the comment management interface
03. projects
Things I've Built
Anita Jobs Alert Bot
Autonomous production-grade job scraping bot built with Python and asyncio. Scrapes 23 job categories every 30 minutes, delivers real-time Telegram alerts, and supports conversational LLM-driven job queries via Claude Sonnet through the Anthropic API.
- 2,000+ listings surfaced since deployment
- Claude Sonnet tool-use pipeline for dynamic filtering
- 100+ company URL resolver with async parallel scraping
GitOps Deployment Pipeline
End-to-end GitOps pipeline using ArgoCD and Minikube for continuous delivery with self-healing on every Git push to main. Containerized a Python/Flask app via Docker, deployed on Kubernetes with zero-touch Git-to-cluster reconciliation.
- Sub-minute sync latency, zero-touch reconciliation
- Self-healing clusters with automated pruning
- Versioned DockerHub images via CI/CD
Search Engine
Full search engine backend indexing 15+ documents with sub-150ms query response using TF-IDF scoring and inverted index data structures to rank and retrieve relevant documents from a corpus.
- Sub-150ms query response time
- TF-IDF relevance ranking
- Inverted index for fast lookups
04. startup
Gnome
Winner — Harriet Stephenson Business Plan Competition · $20,000 Grand Prize
AI-Powered Plant Care Platform
Gnome pairs a 5-sensor hardware kit (soil moisture, light, temperature, humidity, root-zone) with a conversational AI assistant that turns real-time IoT data into personalized plant care guidance. As AI & Product Integration Lead, I architected the RAG pipeline, conversational intelligence layer, and multi-tier cloud infrastructure powering Gnome's recommendations.
My Role
- ▹Architected the async RAG pipeline: S3 Parquet → Athena batch analytics → Amazon Titan embeddings → Pinecone vectorstore
- ▹Built the live chat orchestration Lambda aggregating ElastiCache session context, Pinecone semantic memory, and RDS live sensor metrics
- ▹Implemented serverless backend with AWS Lambda, API Gateway, SQS, and Docker Lambda Layers to resolve cross-OS Python dependency issues
IoT Sensor Stack
ESP32 microcontroller tracking soil moisture, light, temperature, humidity, and root-zone conditions — syncing hourly to the cloud via API Gateway + SQS.
Serverless Architecture
AWS Lambda backend with SQS decoupling, RDS persistent storage, and ElastiCache (Redis) for sub-millisecond session caching during active user interactions.
AI-Powered RAG Pipeline
Async Athena + S3 Parquet analytics, vectorized with Amazon Titan into Pinecone for contextual, history-aware plant care responses.
Tech Stack
5
Sensor Types
$1.82M
5-yr Revenue Projection
35.6M
Addressable Households
$20K
Prize Won
05. skills
Tech Stack
Languages
Frameworks & Libraries
AI / ML
Databases & Cloud
DevOps & Tools
05. education
Education
University of Washington
Seattle, WA
B.S. Informatics — Data Science & Software Development
Relevant Coursework
07. passions
Beyond the Code
Music
Lead soloist for UWAwaaz, a Western and Bollywood fusion a cappella group at UW. We perform at university events, competitions, and community shows across the US.
Formula 1
Big F1 fan and a proud Williams supporter through every season. Love the engineering, strategy, and the chaos that comes with being a Williams fan.
Sports
Following four clubs across three sports. From the Pacific Northwest to the Bundesliga, Premier League, and La Liga — always watching.
Golf
Love getting out on the course whenever I can. Golf is the perfect mix of focus, patience, and the occasional miraculous shot.
Building Software
Outside of work and school I genuinely enjoy building things — side projects, bots, tools, and whatever interesting problem catches my eye. The best hobby is one that doubles as a skill.