Senior Full Stack Engineer - Applied AI
First seen by Alion on Oct 7, 2026.
Fifteen professional sports teams use our product every day. You would be the engineer who owns it.
Pixology is an AI-native platform for the sponsorship economy. Sports teams use us to turn their sponsorship inventory and a brand's objectives into visual concepts, tailored pitches and the films that win deals. Work that used to take a designer two weeks now takes hours.
We are live with fifteen teams across nine leagues, including clubs in the NFL, MLB, NBA, NHL and MLS. Three founders, out of MIT. You would be our first engineering hire in India.
THE HONEST VERSION OF THIS JOB
Our CTO built this platform largely with AI coding tools. It works. Real customers, real usage, real revenue. It also has the shape you would expect from that: patterns that differ between files, test coverage that is thin in places, and decisions nobody wrote down.
You are not joining to maintain somebody's tidy codebase. You are joining to take ownership of a working product that needs a proper engineer's hands on it, and then to build the next two years of it.
If that reads as a problem, this is not your role. If you have inherited a system before and found the archaeology satisfying, keep reading.
WHAT YOU WILL OWN
The whole stack. Frontend, backend services, database, API integrations. There is nobody else to hand a ticket to.
Applied AI. Generative models that turn sponsorship inventory and brand context into visual concepts and pitch material. Production AI, not a demo.
Evaluation infrastructure. Harnesses and regression tests that let us compare models, validate prompt changes, and track output quality, latency and inference cost. This barely exists today and it matters more every month.
The media pipeline. Timeline-based video editing. Clip trimming and sequencing, transitions, audio synchronisation, rendering and export. Genuinely hard engineering, and the part customers notice most.
Shipping. Design, build, test, deploy. Customer bugs come to you.
WHAT WE ARE LOOKING FOR
Around five years building production software, with real ownership of something live.
Strong across the stack, with a view on where logic belongs rather than just the ability to write both halves.
You have inherited a codebase you did not write and made it better without rewriting it.
Working knowledge of LLM APIs and the practical tradeoffs: cost, latency, reliability, evaluation.
You debug well. Given an unfamiliar system and a failing case, you find the cause rather than the symptom.
Nice to have: media processing (FFmpeg, rendering pipelines), computer vision, evaluation tooling, early-stage experience, any interest in sport.
COMPENSATION AND SETUP
40 to 50 LPA, all cash. No equity, and we would rather say so now than in your third interview. We have taken the view that we would rather pay you properly today than hand you a number on a slide that may be worth nothing in four years.
Hybrid. At least one day a week together in our Basavanagudi office, the rest is yours. If you live in South Bengaluru, you never touch the Outer Ring Road.
HOW TO APPLY
No CV screen. Answer these two questions and we will read every one.
- Describe a codebase you inherited that you did not respect. What did you change first, and what did you deliberately leave alone?
- Tell us about a production bug where the cause was not where you first looked. How did you find it?
Skills
React.js, TypeScript, NodeJS (Node.js), Python, PostgreSQL, Generative AI, Large Language Models (LLM), RESTful APIs, FFmpeg, Amazon Web Services (AWS), Docker