Production
Amogh Matt
Role
Engineering Manager
Company
BMAT, Barcelona
Running time
2015 to now

I help build and run the software that tracks where music is played and who gets paid for it.

Since 2018 I've worked on music reporting at BMAT, first as an engineer, then as Product Owner, and now as an Engineering Manager in Rights & Royalties. Before that I ran infrastructure in Bangalore and studied sound and music computing in London.

From play to pay

The steps between a song being played and someone getting paid for it. I've worked on every one of them, and each links to the work below.

  1. PlayA track airs on TV, streams, or fills a venue.
  2. DetectAudio fingerprinting picks the music out of the signal.
  3. MatchRecordings are linked to works and the people behind them.
  4. ReportCue sheets and usage reports go to collecting societies.
  5. Claim & payUsage becomes claims, then distributions to rights holders.

Selected work

Selected work as numbered cues, oldest first
CueInOutTitleStageRoleNotes
0120152018Before BMATBefore music rightsDevOps engineer, then researcherData-over-sound research at Chirp (EUSIPCO 2019), and infrastructure DevOps at SapientNitro.
0220182024Reportal & CUEDDetect, Match, ReportEngineer, then Product OwnerOne platform, two products: Reportal reports broadcasters' music use to collecting societies, CUED builds cue sheets for production companies. From the first broadcaster system to a multi-tenant platform on Kubernetes.
032023nowInnovation projectsPlay, DetectTechnical lead, Product OwnerMusic recognition and insights for venues, an audit system for generative AI music engines, and automatic cue sheets for audiovisual productions.
0420242026Leading Reportal & CUEDReportEngineering Manager, Reporting, 6 peopleSprints, retros, a tech debt vote, a cloud clean-up and incident response, while unifying per-client services.
052026nowDSP Processing & ReportingClaim & payEngineering Manager, Rights & Royalties, 10 peopleTwo-week sprints, continuous deployment, a dev-on-duty support rotation and an incident process.

How I work

Before the sprint

  • Writing things down as we go

    I keep runbooks and decision notes up to date while the work happens, so nobody has to remember how something works or why we chose it.

  • Priorities before planning

    I agree priorities with Operations and Business before a sprint starts.

During the work

  • The team picks the tech debt

    When there's time for clean-up, the team votes on what slows them down most. They know the code best, and work they chose tends to get done.

  • Knowing what we run

    Every service, database and bucket has an owner.

When things go wrong

  • One person leads an incident

    When something breaks, the dev on duty leads and the rest of us help. We steady things first, and the proper fix goes into the next sprint.

  • Retros without blame

    Every couple of weeks we look at what went well and what didn't. When something goes wrong, we look at what in the process allowed it.

Credits

  1. 2026 to nowEngineering Manager, Rights & Royalties, BMAT
  2. 2024 to 2026Engineering Manager, Reporting, BMAT
  3. 2022 to 2024Product Owner, Reportal, BMAT
  4. 2018 to 2022Software Engineer, TV & AV, BMAT
  5. 2018Research Technician, Chirp, London
  6. 2017 to 2018MSc Sound and Music Computing, Queen Mary University of London
  7. 2015 to 2017Infrastructure DevOps, SapientNitro, Bangalore
  8. 2010 to 2017Lead guitarist, 1Fret Away, Bangalore

Writing

All writing