Skip to main content
The Semilattice MCP server lets you simulate how real users react to product changes, concepts, and decisions. Use it for discovery, concept validation, feature prioritisation, message testing, conversion flow optimisation, UX comprehension, and usability testing.

​
Concepts

AI representations of specific groups of people, built from real surveys and qualitative sources. Every user model is measured against human ground truth with transparent accuracy metrics. Pick from 50+ ready-made user models (UK/US adults, UK Consumer Finance segments, B2B software buyers, tech audiences) or work with us to build a custom user model from your own data. Read more about user models →
Simulations predict specific kinds of user behaviour, focused on a decision-making question or goal. A simulation might be focused on validating a new product concept, discovering user frustrations around a particular problem, testing lines of copy, or even checking if users can convert through a sign-up flow.
Two modes cover the breadth of product and marketing cycles. User Research simulates your users answering questions. Use it to explore preferences, intents, and attitudes in conceptual space. User Journey mode simulates users interacting with user interfaces. Use it for conversion flow testing, UX discoverability and comprehension, or UI problem discovery.

​
How it works

1

Plan

Your assistant picks a user model, picks a mode and a goal, drafts the simulation, and refines the plan with you.Tools: list_user_models, list_simulation_schema, create_simulation, update_simulation_plan
2

Run

Your assistant kicks off the simulation and polls until it’s done. Simulations can take anywhere from a few minutes to half an hour.Tools: run_simulation, get_simulation (poll)
3

Interpret

Your assistant pulls the finished simulation and folds the results into the next step of your workflow.Tools: get_simulation, list_simulations

​
Tools

​
Plan

  • list_user_models: browse available user models with their accuracy metrics.
  • list_simulation_schema: discover modes (User Research, User Journey) and the goals available within each.
  • create_simulation: draft a simulation against a chosen user model and goal.
  • update_simulation_plan: refine the simulation’s plan before running.

​
Run

  • run_simulation: start a planned simulation.

​
Interpret

  • get_simulation: fetch a simulation’s status, plan, and results once complete.
  • list_simulations: browse past simulations to build on.

​
Connect

Connect Semilattice MCP

Set up the Semilattice MCP server in ChatGPT, Claude Code, Cursor, or any other client that speaks remote MCP over Streamable HTTP with OAuth.