Product Roadmap
Last updated: March 16, 2026
Our product roadmap focuses on reliability, observability, and developer experience for mission-critical desktop automations.
Shipped
- 93–96% click accuracy with reflection agent that verifies and self-corrects
- Handles UI changes, unexpected dialogs, and vendor updates automatically
- Human-in-the-loop fallback with full context when recovery fails
- Full video replay of every automation run
- Screenshots at each step with execution context
- Slack failure notifications with direct debugging links
- Error pattern analysis via AI assistant
- Cloud, on-premise, or Citrix/RDP deployment
- Containerized on-prem. No data leaves your perimeter.
- Horizontal scaling across Windows VM pools
- SOC 2 Type II and HIPAA compliance with BAA
Coming Soon
Expose workspaces as MCP (Model Context Protocol) servers, enabling AI coding agents and other tools to trigger and interact with desktop automations directly.
Automated testing and validation of desktop automations using AI to view screenshots of input/output per workflow step, catching regressions before they reach production.
- Git integration: Turn workspaces into git repos for editing workflows in VS Code
- OpenAPI spec generation: Auto-generate OpenAPI specs for workflows (input/output schemas)
- Diff viewer: Visual comparison of workflow changes before deployment
Research Phase
Fully autonomous agentic workflows that can reason about and execute multi-step processes without predefined scripts.
Why We're Taking Our Time
Current LLM automation approaches are not yet reliable enough for mission-critical production workloads. We focus on deterministic self-healing agents with AI-assisted debugging until agentic approaches can match the reliability standards our customers need.
Our approach: We'll only release APA when it can deliver the same 99%+ reliability that customers expect from Minicor today.
Have feedback on our roadmap or feature requests? Contact our team to share your input.