Matchday production workspace

A reviewable matchday content set for production teams, from briefing notes to post-match drafts.

For

Broadcast, production, and sports media teams.

Input to output

01Fixtures and lineups
02Approved news and media
03Editorial and brand rules
OUTExample production output

Example production output

Producer briefing

Lineup change: source-linked context and talking points

Draft lower-third and social copy for review

Teams typically build this example with Factory and run it in Machina Studio alongside their other projects and workflows.

Explore this build in Factory Plan a guided build

How it works

1
Editorial Guidelines
Upload commentary templates and branding guidelines.
2
Sports Data Synchronization
Process live match data and identify key moments.
3
Content Generation
Create dynamic commentary and analysis using Machina AI.
4
Video Production
Prepare supported video segments when the required media tools and rights are configured.
5
Media Management
Prepare approved media assets for the configured content workflow.

Example components, sources, and tools

Setup notes

Sportradar
HeyGen
Machina DBMachina component
Machina AIMachina component
VideoDB

Machina components are shown separately from external source and tool connections. External availability depends on the configured build and source rights.

The Matchday production workspace is an example agent that produces a broadcaster's match-day content set — previews, recaps, run-sheet material, graphics copy and social posts — from fixtures, form, squads and confirmed lineups. It runs on the fixture calendar rather than on a prompt, so configured fixtures beyond the marquee ones can get the same treatment.

What the Matchday production workspace produces for a fixture

One agent run covers a fixture end to end, drawing every surface from the same grounded source. Shared sourcing keeps the script, the graphic and the post aligned, and a cross-surface check catches any number that drifts before review.

A preview once the fixture is confirmed, rewritten when the lineup lands.
Run-sheet material: talking points, stat blocks and storylines for the producer.
A recap drafted after the final whistle, in the languages you configure.
Per-platform social copy tied to the same facts as the broadcast script.
Graphics and lower-third copy for the same moments.

Where it fits in an existing production process

It does not replace the desk; it changes what the desk starts from. Producers move from assembling research to reviewing a draft that already carries the club data, the form and the storylines — which is a different job with a different time cost.

The approval step is the part worth designing carefully. Review has to complete inside the window between lineup confirmation and kickoff, so the practical pattern is human sign-off on anything going to air and spot checks on lower-risk surfaces such as long-tail recaps.

What it needs from you

Three inputs, and the third is the one teams underestimate: a data source for fixtures and lineups, the languages and platforms you publish to, and your brand rules — tone, banned phrasing, sponsor and compliance constraints — expressed as rules the agent applies at generation rather than notes a reviewer applies afterwards.

Full guideAI Content Automation for Broadcasters: What to Automate FirstThe production constraints that decide whether a deployment survives a live gallery, and how to scope the first one.

Last updated: 2026-07-30

Frequently Asked Questions

What is the Matchday production workspace?

An example agent that produces a broadcaster's match-day content set — previews, recaps, run-sheet material, graphics copy and social posts — from fixtures, form, squads and confirmed lineups. It runs per fixture on the calendar rather than per prompt, so the configured fixtures get covered, not only the marquee ones. Teams build it with Factory and operate it in Machina Studio.

Does the output go straight to air?

Not by default. Anything going to air should carry human sign-off, and the approval step has to fit between lineup confirmation and kickoff. Lower-risk surfaces such as long-tail recaps usually move to spot checks once the desk trusts the voice.

Can it publish in more than one language?

Yes. The same run localizes per market, with brand rules applied per language rather than re-implemented by a separate desk. For broadcasters with an international audience this is normally where the workload difference shows up first.

What happens when the lineup arrives late or a feed fails?

The behaviour is configured rather than left to the model: hold, degrade to a shorter format that does not depend on the missing field, or flag a human. That decision matters more in production than generation quality, because late lineups are routine, not exceptional.