ASK AXON AI
Connect Your AI Tool Directly To Your Performance Data
Your own AI tool, querying your warehouse directly via the Ask Axon AI layer.
Pre-defined measures and sport-specific agents built in.
Accuracy regularly reviewed and measured.
TWO WAYS TO USE ASK AXON
Two Access Points.
Same Data Foundation.
Analysts make advanced queries and develop reporting using their own AI tool, Claude, Gemini or ChatGPT via an MCP. Coaches can ask questions directly in Axon HQ, same AI layer, same data underneath.
ASK AXON IN AXON HQ
For anyone on the team who wants a quick answer.
Ask Axon sits inside Axon HQ with its own chat window for quick one off questions.
Example questions
Which players have a wellness flag and an elevated ACWR going into the weekend?
Who has missed more than 20% of training load in the last 14 days?
Show me our squad availability summary for Saturday.
ASK AXON VIA MCP
For performance analysts. Data leads. Sports scientists building repeatable, complex workflows.
Connect Claude, Gemini, ChatGPT, or any AI tool directly to your Axon data warehouse. Ask complex questions across multiple datasets. Build match reports, trend analyses, and flagging workflows that can run consistently every week.
Example workflows
Daily Readiness report pulling GPS, Force sensor, wellness and wearable data.
Season trend analysis across a squad. Load, recovery, physical output, game minutes
Trend, match, recruitment, referee or competitor analysis of game data
WHAT IS AN MCP FOR SPORTS TEAMS
AI Answers Are Only As Good As The Data Underneath.
Most organisations trying to connect AI to their performance data are still collating csv, excel and xml files, then uploading them to the cloud. This process has a number of pitfalls.
The data is spread across six or more source systems with no coherent way to connect them.
The AI doesn't know how to analyse the data, so the analyst has to explain it each time
AI tools struggle to correctly analyse large amounts of data in files due to context windows that limit the amount of rows it can consume at one time.
Ask Axon uses a technology called MCP (Model Context Protocol). The MCP stores ‘context’ so that when an analyst asks a question, or builds a workflow, the MCP helps to translate the question.
That translation could be a query on a database for a single question, or the creation of an entire report across multiple data sources.
The Axon MCP contains over 2,000 pre-defined measures and definitions, as well as a series of agents that run every time a question is asked.
These agents are specialists in their domain, so the system knows what you mean when you ask.
Users
AI tool or Axon HQ
MCP Server
Data Warehouse
HOW WE MEASURE ACCURACY
How Can You Trust The Answers?
Being able to trust the responses given by Ask Axon is the most important part of using the system.
To be sure the MCP is giving accurate responses Ask Axon evaluates the performance of these responses.
We regularly run anonymised evaluations of whether an agent understood the question, queried the right data, retrieved sufficient evidence, and stayed faithful to that evidence.
More information on how we measure accuracy.
THE SECURITY MODEL
Secure Access To Your Data.
When an analyst uses a cloud AI tool they often upload raw athlete data to the cloud. On top of the reduced effectiveness, this adds security risk to the data.
When using the MCP directly on your data warehouse, you have access to all of your data in a secure location while only the required response is returned.
Using a desktop tool, such as VS Code, with your AI tool allows analysts to store everything they work with on their desktop.
Table names, column names, data types. Your AI tool sees the schema so it knows what to query.
Your AI writes a SQL query. The query executes on your warehouse.
Snowflake sends back just the data that is required to answer the question.
SEE ASK AXON LIVE
See This In Action With Live Queries.
Book a 30-minute call with an Axon analyst or set up a trial with your own data.