An AI layer that speaks your game.
Assistant, automation, and a virtual workforce you can hire. Managed and sold as labor, priced per agent-hour.
Six things the AI pillar delivers.
Ad-hoc data queries
Ask questions of game data by voice or text — ARPPU per cohort, retention curves, offer performance.
AI-authored admin screens
Render contextual screens per entity on open. Every schema becomes a workable surface without a form builder.
Campaigns without clicks
Describe a live-ops campaign in plain language; the platform compiles it into scenarios for review.
Sandboxed before they run
Every generated scenario lands in a sandbox first. A human approves before it touches the live game.
Virtual agents as roles
Assign an agent to LiveOps, QA, CS, or marketing. It works inside your studio, with your data, on a role you defined.
Billed as labor
Assistant and automation included in the Platform tier. Virtual agents priced in tokens or agent-hours, not seats.
AI fits three kinds of operators.
Buys labor, not a tool
You look at a studio of 15 and ask which of those roles an AI can carry. This pillar answers the question with a line item.
Automates the weekly grind
Campaigns, event rotations, offer tuning — composed by voice, reviewed before they run.
Queries the game, not a BI backlog
Ask for "whales who haven't paid in 14 days" and get a list — with confirmation before any action.
Anchor story forthcoming.
In our studio there are 15 AI agents and 3 humans. A studio CEO walks through hiring agents as employees — LiveOps, QA, marketing, support — and the operating model that makes it work. Sample journey commissioned; lands here when the game does.
Walk through early access →AI composes three modules into a horizontal layer.
- In the platform →PlayServ AIAssistant, automation, and virtual agents. The three-tier progression from ad-hoc queries to campaign composition to role-level labor.
- In the platform →Adaptive AdminAI-generated admin screens per entity, per context. The render surface where automation and agent output lands.
- Open spec →Forge · Extensible DataTyped schemas and relationships — the context model agents reason over. Every entity is a first-class object for automation.
Assistant and automation included. Virtual agents priced as labor.
Assistant and Automation tiers are included in the Platform tier as fair-use. Virtual Agents are billed separately as a new line in tokens or agent-hours — contact sales for enterprise terms.
Ready to hire your first AI agent?
Start with the assistant — free. Scale into a virtual workforce when the game is live.