Computer Science & Statistics · UC Berkeley

Shri Charan Sriram.

I build machine learning models and full-stack products, with a focus on data, markets, and measurable results.

Areas of focus

  • Machine learning
  • Statistical modeling
  • Financial markets

01 / Selected work

A few things I've built.

Selected projects across finance, consumer data products, and applied ML research.

02 / The project index

More things. More code.

Projects across machine learning, web development, and hardware. Filter or search below.

Showing 8 of 8 projects

Finance · Simulation

IndexCheck

A React portfolio stress-tester with historical crash replays, a 10,000-path Monte Carlo engine, and an LLM layer that turns drawdowns into plain-English risk explanations.

React / TypeScript / Monte Carlo / LLM API
Campus · Real-time

Berkeley Free Food Finder

Students post free food on campus in real time. No accounts, rate-limited posting, and posts that expire on their own once the event ends.

React / Node.js / TypeScript / SQL
Fragrance · Data viz

Accord

A map of 24,063 perfumes where distance means similarity. Dots sit near the bottles they share ingredients with, so you can search one you already own and find others worth sampling.

React / Vite / Data visualization
Sports analytics

NBA Shot-Make Model

A shot-make probability model for the OKC Thunder analyst project, trained on 426K+ labeled NBA shots. Defender distance and approach time series parsed into features, benchmarked on log-loss against OKC's model.

Python / pandas / scikit-learn
AI agents

Specter

An agent you train by demonstration. It records a workflow once and replays it to automate repetitive tasks. 2nd in Technical Achievement out of 60+ teams at the Ara × Stanford Hackathon.

AI agents / Learning from demonstration
Health data · Backend

Clinical Trials, Charted

A backend service that takes a plain-English question about clinical trials and turns it into a chart spec, pulling the data from the ClinicalTrials.gov API.

Backend / NLP / ClinicalTrials.gov API
ML research

AI Dermatology Research

A compression-robust classifier for skin cancer detection. 10,500 dermatoscopic images, 90% accuracy under 70% compression, published in the Journal of Student Research.

Python / TensorFlow / Computer vision
Embedded

Gesture Controller

A gesture-based computer controller built on an LIS3MDL magnetometer. 90% accuracy across 6 gestures. Built during the UCI × GATI fellowship.

Sensors / Embedded / Signal processing

03 / Experience

Where I've worked.

Select a role to see the scope of the work, the tools used, and the results.

  1. 2026

    San Ramon, CA

    Helping students tell their own story.

    • Advise 5+ students on UC Personal Insight Questions and Common App essays.
    • Built a Google Forms intake pipeline and workflow to manage clients end to end.
    • Standardized an editing framework across 25+ essays, swapping generic language for scene-driven detail.
    • Writing
    • Admissions
    • Operations
  2. 2025
  3. 2024
  4. 2024
  5. 2024

04 / About

A little about me.

Computer Science & Statistics.
Focused on ML and markets.

GPA
4.0
Double major
CS + Stats
Graduating
2029

I'm studying Computer Science and Statistics at UC Berkeley. My work sits between machine learning, software engineering, and quantitative analysis.

Outside of coursework, I build and ship my own products, run an admissions consulting practice, and develop trading agents.

I'm especially interested in equity markets and IPO analysis: how companies are valued, priced, and received when they go public.

My toolbox
  • Python
  • NumPy
  • pandas
  • Java
  • TypeScript
  • React
  • Node.js
  • SQL
  • Git

05 / Off the clock

Off the clock.

What I'm into outside of coursework and projects.

Watches

I'm into mechanical watches and the small details on a dial.

grail: Seiko Imperial or a Versace chronograph

Chess

Casual games online between problem sets. Every game is a puzzle I can actually finish.

rating: ~1200 and climbing

DJ'ing & making beats

I make beats in FL Studio and I'm getting into DJ'ing. Mostly house music.

currently on repeat: 925 by Sammy Virji

Spikeball

Pickup games with whoever's around.

home courts: Memorial Glade, Willard Park

Travel

Hard to pick a favorite so far. It's between Costa Rica and Singapore.

next up: Amalfi or New Zealand

Have an idea? A good question?

Let's build something useful.

Open to internships, research, and collaboration.