Srimanth Reddy

Software Engineer at TELUS AI. I build AI infrastructure and data systems.

I'm a Software Engineer II at TELUS AI (formerly Playment, YC W17), working on the infrastructure and data pipelines that power computer vision, LLM, and Physical AI annotation workflows.

Before going full-time here I ran a few ventures — a scooter fleet, a DOOH advertising startup, a food delivery app — and learned most of what I know about operations, hardware, and product by breaking things in production.

Now

Career

TELUS AI · formerly Playment (YC W17)

Software Engineer II · Apr 2026 – Present

AI infrastructure and annotation systems. Current focus is on data pipelines for Physical AI — the tooling required to collect, label, and version the sensor data that trains agents operating in the real world.

Software Engineer I · Aug 2024 – Mar 2026

Full-stack across the annotation platform — frontend tooling, backend services, and the infrastructure supporting computer vision and LLM data pipelines.

Software Engineering Intern · Jul 2023 – Jul 2024

One-year internship during my final year at BITS Pilani. Worked across computer vision annotation workflows, QC pipelines, 3D web rendering, Protobuf/gRPC, and microservices. Converted to full-time on graduation.

Ventures

Xpress Bikes · 2022–2023

Started a scooter rental fleet in college — 2 scooters, grew to 20, serving students and commuters around campus. First real exposure to fleet operations, battery degradation curves, and what it costs to keep physical hardware running.

Adcel Media · 2023–2024

Digital out-of-home advertising on a budget: Raspberry Pi, LED panels, custom networking. Built the display hardware and software stack from scratch. Ran into manufacturing tolerances, supply chain lead times, and the economics of hardware businesses. Didn't ship at scale — but I learned more about embedded systems and networking in six months than I had in years of coursework.

Qzin · 2024–2025

10-minute food delivery. Built the mobile apps, backend, delivery dispatch, and ops tooling myself. Got to 1,800 users and 20–25 orders a day at ₹200 AOV. Shut down when unit economics didn't converge. The hard lesson: logistics density math is unforgiving.

Interests

There's a thread through everything I've worked on: systems that touch the physical world. A scooter fleet taught me battery systems. A display startup taught me embedded hardware. Annotation infrastructure taught me what it takes to train models on real sensor data. Now I'm following that thread into Physical AI — the infrastructure, data systems, and tooling needed to train agents that work in the real world.

Reading

A small collection of books and ideas that have shaped my thinking.