All projects
07

Sports streaming and scouting

JUU

A platform for local-league streaming, player discovery, and AI-generated highlights of goals and important match events.

Client

JUU team

Location

Private engagement

Engagement

Sports platform engineering

Delivery

Production product

[ 01 ]

The challenge

Promising players in local leagues often lack the coverage and reusable footage that help scouts notice them. A full match stream alone does not solve the problem if the decisive moments remain buried in a long recording.

JUU needed to connect three demanding workflows: live coverage, player visibility, and automated extraction of goals and significant match events.

[ 02 ]

How we handled it

[ 01 ]

Connect players, leagues, and matches

The information model treated profiles, teams, fixtures, streams, and highlights as related parts of the same scouting record.

[ 02 ]

Build a dependable media pipeline

Streaming and recorded media were organized so match footage could move into processing without losing its league and player context.

[ 03 ]

Make AI output reviewable

Detected goals and match events were shaped into clips that could be checked, attributed, and used in player discovery.

[ 03 ]

What went right

The platform solved distribution and discovery as one product rather than two disconnected tools.

Structured match context made generated highlights more useful to scouts.

The workflow recognized that AI-detected events still benefit from clear review and attribution.

[ 04 ]

Outcome

OUTCOME 01

Local leagues gained a way to broadcast matches while creating reusable moments from the footage.

OUTCOME 02

Scouts could move from a highlighted event to the player and match context behind it.

What we delivered

  • League and player profiles
  • Live-stream workflow
  • AI highlight pipeline
  • Scouting experience

Behind the scenes

For readers who want the technical detail, these are the specialist building blocks supporting the experience. Clients do not need to select or manage them.

Amazon IVS low-latency videoRTMPS/HLS deliveryFFmpeg transcodingVision event detectionTemporal clip generationPlayer metadata graph