AI assistants
Let your own assistant ask LoadGen.
The LoadGen Platform ships an MCP server. Point the assistant your organisation already uses at it, and the questions you would otherwise click through become questions you can ask.
The capability
An MCP server, built into the Platform
The LoadGen Platform exposes its own operations over the Model Context Protocol, so an assistant your organisation already runs can talk to it directly.
Over 650 tools
The operations the Platform already performs are published as tools, rather than a separate, smaller API written for assistants.
Your assistant, your conversation
Ask about a run, a schedule or a result in the assistant you already use, instead of learning a second interface for the same questions.
The security model
Off by default
The MCP server is not enabled for you. Nothing is reachable until an administrator turns it on deliberately.
An administrator holds the key
Access is granted by a key an administrator issues, and that administrator can rotate it.
The Platform’s own permissions still apply
An assistant reaches the Platform through the same permission model as any other caller. Turning the server on does not widen what an account may do.
Auditable like the rest of the Platform
Calls arrive through the Platform, so they sit alongside its other recorded activity rather than beside it.
Two different things
This is not the AI builder.
LoadGen has AI in two places, pointing in opposite directions. It is worth being clear about which one you are asking for.
The AI builder, inside Studio
You describe a journey and Studio builds the workload on the real application. The AI is producing something for you.
Studio AI builderThe MCP server, on the Platform
Your assistant reaches into the Platform to ask about what is already there. The AI is not LoadGen's; it is yours, and LoadGen is the thing it is talking to.
This pageWhich assistant you run, and which model is behind it, is your choice. LoadGen does not require a particular one and does not supply it.
Ask us what your assistant could do with it.
We will walk through the permission model, show what the tools cover, and be straight about where an assistant helps and where the Platform's own screens are still faster.
