Inference-time reasoning harness for AI agents that cuts inference costs by 60% to 85%
Stage
Not given
Startup
Founded
Not given
2025
Based in
DE
US
Pricing model
Freemium
Not given
Pricing
Free: €0/month with 25 on-screen requests per month, 2 projects, and 5 connected tools. New users get a 14-day Pro trial. Pro: €19/month or €199/year for unlimited on-screen requests, unlimited projects, unlimited integrations, knowledge graph memory, and priority support. AI provider charges are se
Not given
Key features
Computer use: clicks, types and fills forms in apps
Voice dictation and screen marking
Memory that learns user details and reuses them
Parallel agents running simultaneously
On-screen guidance showing where to click
Integrations with Gmail, Slack, Notion, GitHub and more
Task board to track running, done and waiting tasks
Projects as workspace homes with tasks, chats and results
Skills: save and reuse how you do a job
Works with Claude Code, Codex and OpenCode
Sub-2ms verification latency with deterministic rule matching
311 pre-built verification rules across four protection domains
Zero Data Retention policy with no prompt logging or training
Python SDK, MCP server, REST API, and n8n integration support
Fail-open design ensures production workflows never stall
Adaptive mode for synthesizing tailored scaffolds for novel tasks
What makes it different
Works directly on your screen in your apps instead of forcing you to a chat window
Visual input through pointing, drawing, and marked screenshots instead of text prompts
Full user control: asks before acting and shows cursor while working
Remembers user details and context across tasks
Voice-based task dispatch from any app you're already using
Enables cost-efficient models to match frontier accuracy, saving 60-85% on AI inference bills
Injects structured invariants before token generation to prune invalid failure paths upfront
Based on 46 years of maritime command and deep-sea research operational discipline