Athlete's AI in Collaborative Coding

Athleteai Collaborative Coding

Since mid-2026, AthleteAI data can be ingested from a secondary computer through collaborative coding. The primary (remote) coder carries on with their match-day process, while a secondary coder runs AthleteAI to bring through extra ball-tracking variables in real time for match-day analysis and insights.

In this article

1. Match shell setup
2. Enable collaborative coding ingestion
3. Process the video with AthleteAI
4. Ingest through collaborative coding
5. Review, remediate and analyse
Where to go next

1. Match shell setup

Before the match, tick Write requests and Read responses in the match shell, and choose the camera angles you want AthleteAI to process.

Set this up before the match. Once play starts, you can only change these settings during breaks in play.

2. Enable collaborative coding ingestion

Before the first ball, check that collaborative coding is enabled in the coder setup. This lets data collected through collaborative coding flow through to the primary coding computer and on to the stats server.

Collaborative coding enabled in the coder setup

3. Process the video with AthleteAI

On the secondary computer, make sure Receiver is running, so write requests come through during the match, and keep collaborative coding open.

How you run AthleteAI from here depends on the use case:

Workflow	When to use it
Video Cutting Reviewer	The standard live workflow (see the Video Cutting Reviewer documentation). Make sure AthleteAI only looks for the views you've requested in the script. Keep it away from broadcast views or anything else that could affect calibration and processing time.
Direct 50 FPS processing	The primary coding computer is capturing at 50 FPS but can't run AthleteAI itself. The secondary computer runs AthleteAI directly from the MatchID and processes the video in the folder.
Long video mode	Several cameras are set up for post-match processing. The Video Cutting Reviewer is still recommended, because it follows the naming conventions and has lead and lag times built in, but long video mode is available post-match.

Choosing a processing mode in the CricketAI app — Processing modes in the CricketAI app

4. Ingest through collaborative coding

Once AthleteAI has processed the video, bring the data through Analyzer and collaborative coding.

Open Analyzer and collaborative coding.
Select the match and the batting innings.
When the collaborative coding control appears, work through the usual processes, then enable the Athlete's AI coding option.
The AthleteAI log starts processing the XML files. Keep it open throughout.

The collaborative coding control with the AthleteAI option enabled — The AthleteAI option in the collaborative coding control

5. Review, remediate and analyse

Once the data is in the match, review, remediate and analyse it using whichever workflow suits you.

Tip: the bulk edit ball-tracking screen in the coding setup is the easiest way to find and fix outliers.

Where to go next

Collaborative Coding: how collaborative coding works in general.
Running AthletesAI after a match: processing match video that's already been coded.
AthleteAI installation: getting AthleteAI onto the secondary computer.