Roopa Srinivas

Roopa
Srinivas

I build software where technology meets human impact: control software for lab automation, full-stack apps people actually use, and better AI through careful evaluation. Graduating in December 2026 with a B.S. in Electrical Engineering and Computer Science and a minor in Bioengineering.

Builder, thinker,
problem solver.

I'm a fourth-year EECS student at UC Berkeley with a minor in Bioengineering. My interests span AI, full-stack development, and the ways technology can improve people's lives, particularly in healthcare and education.

I've built control software for a lab-automation instrument, helped improve large language models through RLHF evaluation, shipped a donation app on iOS, Android and the web, and led the mechanical design of a 120-pound competition robot.

Outside of engineering, I compete as a dancer on UC Berkeley's Zahanat team, where I've also managed logistics for 20+ members at national competitions.

Dec 2026

Expected graduation

Internships and fellowships

10,000+

Clinical trial entries analyzed with LLMs

120 lb

Competition robot designed and built

Where I've worked
and what I've done.

Jun – Sep 2026

Software Engineering Intern

Precision Discovery Systems

Contributed to the development of a life-science instrumentation platform for autonomous labs, working hands-on across embedded systems, motion control, and full-stack software.
Worked on system mechanical performance and real-time control, integrating computer vision to support automated operation.
Developed and maintained control software using Python, FastAPI, and React, enabling operator-facing workflows for the instrument.

Dec 2025 – Present
Current

AI/ML Engineer

Handshake AI Fellowship

Evaluated and ranked LLM-generated responses across complex subject-specific prompts to improve reasoning quality via RLHF.
Audited multi-turn dialogues for logical consistency and instruction following, contributing feedback to next-generation model training.
Applied EECS and biology domain knowledge to assess model accuracy on technical problem-solving tasks.

Jun – Aug 2025

Software Engineering Intern

Infera

Built data pipeline integrations using REST APIs and FastAPI to collect, store, and retrieve structured biological metadata across long-running experiment workflows.
Developed frontend interfaces in React and TypeScript enabling end users to query and visualize structured experiment data across projects.
Wrote unit tests with Jest to bring coverage to 70%+ on core data workflows, catching validation edge cases before reaching researchers.

Sep – Dec 2023

Project Intern

PainVox

Analyzed 10,000+ entries from Stanford clinical trials using LLMs, surfacing critical patterns and outliers to support clinical research decision-making.
Synthesized data findings into research summaries supporting grant applications worth $500K–$2M.

Things I've built.

Mobile & Web App

Boost Your Community

A donation app connecting neighbors with homeless shelters in real time, grown from a prototype into a full iOS, Android and web app. Shelters post specific, time-boxed needs, and donors pledge exactly what they can bring.

Live needs map with a 1–50 mile travel radius
Pledges capped at what's still needed, in the database
QR code check-in at drop-off
Organization dashboards and repeating needs
Badges, leaderboard and community heat map
Row level security and required CI checks

React NativeExpoTypeScriptSupabasePostGISGitHub Actions

Project page & live demo Open the app Code

Data Visualization · AP CS Capstone, rebuilt 2026

COVID Vaccine Tracker

An interactive map of COVID-19 vaccinations, cases and deaths in every state from 2020 to 2023. It started as a three-person Java app for our 2021 AP Computer Science capstone; I rebuilt it as a web dashboard around the features our original design doc never got to.

Week-by-week timeline you can scrub or play
Map colored by vaccinations, cases or deaths
A trend forecast checked against what happened
Python pipeline over NYT and Our World in Data

D3.jsJavaScriptPythonJavaProcessing

Open the dashboard Project page Code

Full-Stack Web App

Grocery Store App

A full-stack grocery store app for store managers, with a Python Flask backend and a relational MySQL schema for products, stock levels and customer orders, supporting inventory oversight and order processing.

Python FlaskMySQLHTML/CSS/JS

Project page Code

Financial Analysis Tool

Investment Indexer

A Python tool that ranks the top and bottom-performing investment indexes over 15 years across 1,000+ stocks and funds, using Pandas to generate sortable performance reports over configurable time windows.

PythonPandas

How I lead
outside the codebase.

May 2025 – Apr 2026

Logistics Captain

UC Berkeley Zahanat

Spearheaded logistics for a nationally competitive dance team, coordinating travel, accommodations, and registrations for competitions.
Managed team scheduling and rehearsal calendars for 20+ members, aligning intensive training timelines with national competition deadlines.
Competed as a dancer while simultaneously managing travel and competition logistics, contributing to multiple awards.

Aug – Dec 2025

Instructor, Interviewer & Trainer

Kodely

Led STEM workshops for elementary students and trained incoming instructors on curriculum delivery, safety, and operational procedures.

May 2022 – May 2023

Mechanical Lead

Homestead Robotics, Team 670

Led design and fabrication of a 120-pound competition robot over an 8-week build season using CAD and CNC software.

What I work with.

Languages

Python
Java
TypeScript
JavaScript
SQL
C++
HTML / CSS

Tools & Frameworks

React
React Native / Expo
FastAPI
Flask
Supabase
PostgreSQL
MySQL
Jest
Git
GitHub Actions
Pandas
NumPy
Figma
Linear

Relevant Coursework

Introduction to Artificial Intelligence
Machine Learning for Computational Biology
Foundations of Computer Graphics
Structure and Interpretation of Computer Systems
Data Abstraction and Structures
Instrumentation in Biology and Medicine
Physiology

RoopaSrinivas

Builder, thinker,problem solver.

Where I've workedand what I've done.