LIGHTS OUT AND AWAY WE GO

Hi, I'm Building the future

Full-Stack Developer crafting scalable AI-powered systems with FastAPI, Next.js, and production-grade ML tooling. I design APIs that scale, ship interfaces that feel inevitable, and turn models into shipped products.

> build --with
View Projects Download Resume Contact
Portrait of Varshith Rajarapu
FastAPI
LLM + RAG
PostgreSQL
Next.js
SCROLL
FastAPI· Next.js· React· PostgreSQL· Python· TypeScript· LangChain· FAISS· Docker· OpenAI· Tailwind· WebSockets·

01 — About

The engineer behind the systems

I'm a full-stack developer experienced in building scalable web applications and AI-powered systems using FastAPI and Next.js. I'm strong in API design, system architecture, and database integration with a focus on performance, scalability, and real-world usability. I've shipped production-ready applications and optimized workflows using modern development tools.

Currently pursuing a B.S. in Computer and Information Science at the University of Wisconsin–Eau Claire (Expected May 2026), with a certificate in Information Systems.

Education
B.S. Computer & Information Science
UW–Eau Claire · May 2026
Based in
Eau Claire, Wisconsin
Open to remote / relocation
Focus
AI · APIs · Scalable systems
LLM workflows · RAG · retrieval

02 — Skills

Tools I build with

A pragmatic stack tuned for shipping production-grade AI systems and fast, accessible web apps.

03 — Projects

Selected work

A mix of shipped apps, AI systems, and experiments — chosen for impact and technical depth.

04 — Research

Ad Fraud Detection

Python Pandas Scikit-learn NumPy Matplotlib Jupyter

Detecting fraudulent traffic patterns at scale

University of Wisconsin–Eau Claire · Sep 2024 – Dec 2024

  • Processed and cleaned datasets with 100,000+ records to surface fraudulent traffic patterns.
  • Trained classification & anomaly detection models, improving detection performance by 20–25%.
  • Engineered 15+ features and evaluated models with precision and recall.
  • Produced visual reports to analyze fraud trends and model outputs.
Records processed
100K+
Performance lift
20–25%
Engineered features
15+

05 — Impact

Numbers that matter

A snapshot of scale across research, datasets, and shipped systems.

0
records processed
0
dataset entries
0
concurrent users supported
0
ML accuracy (F1 predictor)

06 — Contact

Let's build something

Got a role, a product idea, or just want to chat about AI systems? Drop a note.

Email
varshithrajarapu@icloud.com
Location
Eau Claire, Wisconsin
Looking for
Full-time · Internships