Football AI agents: one agent for every role at the club
An AI agent does more than answer: it queries the club's data, runs the calculations, prepares documents and carries out tasks, within the limits set by the club.
Apelot configures one agent for each role at a professional football club. Your staff message it from their usual messaging app or a web interface, 24 hours a day.
What is an AI agent for a football club?
An AI agent relies on a language model, like a chat assistant, but it also has tools: connectors to the club's data platforms, files and messaging apps. It picks the tools a request needs, retrieves the data, calculates, then answers or acts.
The difference lies in the data: a general-purpose assistant answers from public knowledge, a club's agent answers from the club's own data (matches, GPS, internal reports, contracts), with the access rights of the person asking.
One agent per role: example requests
Questions asked to the agents. Answers are calculated from the club's data.
Sporting direction
Squad tracking, playing time, youth development, squad strategy.
- How are minutes split among our under-21 players since the start of the season?
- Compare with the three Ligue 2 clubs that give their young players the most minutes.
- Project the wage bill over two seasons with the current contracts.
Coaching staff
Match preparation, reports, team sheet.
- How does Saturday's opponent concede its chances?
- Prepare the team sheet with last week's 4-3-3.
- Generate the pre-match report by position.
Recruitment
Player search, comparisons, reports for the recruitment team.
- Which left-backs under 24 in Ligue 2 and National outperform our starter on crosses and defensive duels?
- When does his contract expire?
- Add him to the shortlist and prepare a report for the recruitment team.
Medical & performance
Training load, alerts, individual monitoring.
- Which players are above their chronic load this week?
- Cross-check GPS load, sleep and wellness questionnaires for the starters.
Logistics & kit
Travel, accommodation, equipment.
- Organise the trip for Saturday's match in Annecy, leaving on Friday after training.
- What equipment do we need next month for the first team and the academy?
Contracts & HR
Expiry dates, drafting, obligations.
- Which contracts expire before the end of the season?
- Remind me of this month's administrative obligations.
How an AI agent works
- Step 1
The request
A staff member messages their agent from WhatsApp, iMessage, Telegram, Slack, Microsoft Teams, Google Chat or a web interface.
- Step 2
The data
The agent queries the sources authorised for that role: data and video platforms, GPS, AMS, spreadsheets, documents.
- Step 3
The calculation
It filters, compares and calculates from the club's data, then prepares the answer: figures, charts or a document.
- Step 4
The action
Before any action visible to others (publishing a team sheet, sending a schedule to the group), the agent asks for approval.
Tools connected to the agents
Sports data and video
Wyscout, Hudl StatsBomb, Opta, SkillCorner, Catapult, STATSports, Kitman Labs, Teamworks AMS, Metrica Sports. Connected through official APIs, using the club's own subscriptions.
Management
Google Drive, Google Sheets, Microsoft 365, Microsoft Excel, Notion, Airtable, HubSpot: documents, spreadsheets, databases and CRM.
Messaging
WhatsApp, iMessage, Telegram, Slack, Microsoft Teams, Google Chat: everyone messages their agent from the app they already use.
Tools without an API
Automated web browsing: the agent uses the tool in a browser, with a dedicated account, in line with the tool's terms of use.
Frequently asked questions
What is the difference between an AI agent and a chatbot?
A chatbot answers from a script or from what it knows. An AI agent has tools: it queries the club's data sources, calculates, prepares documents and carries out tasks, over several steps if needed.
Do the AI agents work with Wyscout, StatsBomb or Opta?
Yes. The connection goes through each platform's official API, using the club's own subscriptions. Excel and PDF exports can also be used.
Can every staff member see all the data?
No. Access is defined per user and per role (RBAC), with you, according to the club's organisation. An agent only uses the sources authorised for the person asking.
Can an AI agent make mistakes?
Yes, like any system built on a language model. That is why answers are calculated from the club's data, accuracy is fine-tuned with your staff during the prototype, and decisions stay with the staff.
Which apps can staff use to talk to an agent?
A messaging app (WhatsApp, iMessage, Telegram, Slack, Microsoft Teams, Google Chat) or a web interface, on phone or computer.
Which AI model do the agents use?
The one the club chooses. Usage is tracked per user, with caps.
Learn more
Get in touch
Tell us about the roles and tools used at your club: we will show you the assistant connected to data similar to yours.