AS
Arrya Sridhar
BITS Pilani, Hyderabad
focus
Machine Learning · Agentic AI · Computer Vision
wants to learn
LLM fine-tuning & MLOps
off-screen
Building autonomous systems, exploring agent architectures
PythonCJavaMATLAB
[ Team 04 — Onboarding Cohort 2026 ]
A team of three from BITS Pilani Hyderabad. We work across machine learning, quantitative systems, and applied engineering. We are here to ship things, break things, and learn out loud.
why the name — JSA stands for the initials of our three members: Jayaditya, Shreyas, and Arrya. No cipher needed — just three people who work better together.
fun fact — our combined caffeine intake during exam week could power a small server rack.
BITS Pilani, Hyderabad
BITS Pilani, Hyderabad
BITS Pilani, Hyderabad
A snapshot of our current coverage. Bars show how many of us actively work in each area — gaps are exactly where we plan to grow during the internship.
Arrya, Jayaditya, Shreyas
Jayaditya, Arrya
Jayaditya, Arrya
Arrya
Shreyas
Shreyas, Arrya
learning together
Jayaditya
Production ML pipelines, agentic frameworks (LangGraph, MCP), modern backend stacks, cloud deployment (AWS, GCP), and the React + TypeScript side of the web we keep avoiding.
Autonomous decision-making under uncertainty, market microstructure & anomaly detection, and the boring-but-hard problem of making research code actually run in production.
A shared sense of how real teams ship software — version control discipline, code review reflexes, writing intent before writing code, and shipping something we'd put our name on without a disclaimer.
Short written update every morning. No 9am meetings, no theater.
Friday — each of us shows working code or a written breakdown. No slides.
Whenever someone is stuck for more than 30 minutes. Switch driver every 25.
At least one approval. Comments are about the code, never the person.
Anything bigger than a small fix gets a short doc before the code.
Default to opening a draft PR with the problem rather than waiting.
We want to move from consumers to contributors this year — starting with good first issues, docs, and small bug fixes before working up to features.