Shauryaveer
Narwal
CS New Grad @ UofM
I spend hours building automation tools just to save a few minutes of manual work. Big fan of building cool things, AI (where it actually makes sense), and clean code.
Featured Projects
A showcase of my recent work and technical expertise.
ShopGPT
AI-powered shopping assistant that aggregates real-time deals and provides personalized recommendations using OpenAI.
Experience
My professional journey and achievements.
Software Engineer Intern (AI/ML)
- Inherited existing log ingestion layer, taking primary responsibility for building processing microservices and AI orchestration
- Built agentic workflow platform using LangGraph state machine with conditional routing to analyze router logs, reducing anomaly detection time from hours to minutes and accelerating issue diagnosis for engineers
- Developed RAG pipeline with PostgreSQL/pgvector and BERT classifier (85% accuracy) for SQL-backed retrieval over router manuals and telemetry, filtering 98% of noisy logs before LLM calls, reducing manual triage by 45%, and lowering projected inference costs by an estimated $15K+/month
- Designed FastAPI microservices deployed on GCP Cloud Run via Docker, using GCP Pub/Sub event queues and automated CI/CD pipelines to support scalable, real-time log processing
Undergraduate Research Intern (Data Engineering)
- Built automated Python pipeline (pandas, scikit-learn, openpyxl) processing spectroscopic Excel data with wavelength ratio analysis and peak detection, cutting processing time by 95% and reducing analysis from hours to minutes
- Applied K-means clustering and statistical analysis (SciPy, NumPy) with interactive data visualizations to accelerate interpretation of biosensor experiments and improve identification of meaningful signal patterns
Software Engineer Intern (Fullstack)
- Collaborated with 6-person development team to enhance global supply chain platform serving 10+ enterprise clients, including Zara and H&M, by implementing 5+ features that streamlined client workflows and platform usability
- Optimized backend operations with Node.js through database query improvements and caching strategies, reducing API response times by 40%; deployed on AWS to improve scalability and responsiveness for high-volume enterprise users
- Implemented mobile order tracking interface using React Native, enabling clients to monitor live shipment status across 500+ daily orders and reducing dependency on manual status checks
About Me
"one must imagine sisyphus happy"
I'm a recent computer science graduate from the University of Michigan (Go Blue!). I've lived in India, the UK, and South Africa, and now I'm based in the U.S.
I'm looking for new grad software engineering roles where I can use AI to solve real-world problems — ideally building things that are actually useful, not just better ad targeting.
Outside of tech, I love lifting weights, hiking, and slowly getting better at snowboarding. One of my goals for 2026 is to read more, and right now I'm working through Dune by Frank Herbert after loving the movies (very excited for the third movie!).
Touching Grass
Gym
I do powerlifting and bodybuilding for fun
Hiking
Love nature and exploring new places
Snowboarding
I'm bad :( Trying to get better at this