Launch HN: Minicor (YC P26) – Windows desktop automations at scale
Recorded: May 26, 2026, 3:01 p.m.
| Original | Summarized |
Minicor | Scalable Desktop AutomationsUse CasesWritingCareersContactSign InBook a DemoGo Live With Any CustomerThe RPA platform for deploying AI into legacy systemsBacked by Y Combinator·SOC 2 Type II·HIPAA CompliantBook a DemoDesktop App - Agent ModeDesktop AppPatient OrdersNew Lab OrderAgent working...Patient: Johnson, Sarah M.Test: Complete Blood CountSubmit OrderSubmit OrderDesktop AppRunningAgent executing workflowErrorUI element movedAdaptingRe-scanning UIRecoveredTask completedIn production withBuild RPA in minutesRadically reduce error rates with self-healing agents.Unlock GrowthMinicor builds and maintains desktop automations so your teams can focus on closing deals and shipping product.01Self-Healing AgentsWhen a UI changes or an unexpected dialog appears, the agent catches it, adapts, and keeps going. A reflection agent verifies every action against what's on screen and self-corrects before the workflow breaks. No more rebuilding scripts every time a vendor ships an update.Writing RPAs isn't hard. Maintenance is the killer.RPA ScriptDesktop AppPatient: Johnson, S.SearchLab ResultsNotesOrdersClick searchWait for resultsMinicor AgentDesktop AppPatient: Johnson, S.SearchLab ResultsNotesOrdersClick searchValidate results02Deploy AnywhereComputer use agents that run on Windows VMs or in the browser. On-premise, cloud, or Citrix. One API call triggers execution.Same architecture running in production at 25,000 patients/day.Minicor Desktop ServiceDesktop Agent v2.1.0C:\Agent> desktop-agent start03Built-in ObservabilityReplay full video recordings of every automation run. Get failure notifications directly in Slack. Debug with screenshots and full execution context instead of guessing.Debug with context, not guesswork.MDesktop WorkflowWorkflow EditorRun// Workflow: Write to Desktop AppStep 1: Open desktop applicationStep 2: Navigate to target screenStep 3: Fill form fieldsStep 4: Submit and verifyExecution History4 runsRun #12 min ago - 12.4sViewRun #215 min ago - 11.8sView!Run #31 hour ago - 3.2sSlack alert sentViewRun #42 hours ago - 12.1sUnder the HoodHow Minicor WorksFrom API call to desktop automation running on a Windows VM. Here's what happens under the hood.1. Your AI system calls Minicor2. Desktop service connects your VMs3. Build and monitor workflows visually4. Scale automatically across VMsOne API call triggers a full desktop workflow. Send the workflow ID and data, get structured JSON back. No SDK, no browser extension.Your AI system sends one requestPOST https://api.minicor.com/workflow/execute |
Minicor is a platform designed for building and running scalable desktop automations utilizing computer use agents, primarily targeting organizations that need to interact with legacy systems that lack modern Application Programming Interfaces. This approach addresses the fundamental challenge that many enterprise systems, such as Electronic Health Records or Enterprise Resource Planning systems, do not offer writable APIs, necessitating interaction with the underlying desktop application as a human would. Minicor solves this by reliably automating these interactions, enabling organizations to transition workflows from manual desktop operation to automated execution in a short timeframe. A core innovation of the platform is its capability to handle dynamic changes through self-healing agents. When user interface elements shift or unexpected dialogs appear, the agent is designed to catch these shifts, verify every action against the current screen state, and self-correct before a workflow breaks. This self-healing mechanism eliminates the need for constant script rebuilding whenever vendors update user interfaces, significantly reducing maintenance overhead, which is often the primary impediment in traditional Robotic Process Automation. This adaptation results in markedly higher accuracy, achieving click accuracy rates of ninety-three to ninety-six percent, substantially outperforming traditional methods that often yield eighty to eighty-five percent accuracy. The platform operates on a distributed, scalable architecture. Computer use agents execute on Windows Virtual Machines or within web browsers, allowing deployment across on-premise infrastructure, the cloud, or Citrix environments. Execution is triggered via a single Application Programming Interface call, demonstrating high scalability, with the system capable of running at levels such as twenty-five thousand patient orders per day in a production setting. The internal operation involves the AI system initiating a request to Minicor, which connects the necessary Virtual Machines, routes the execution, and manages the workflow visually. The process leverages an API call to execute a full desktop workflow, which results in a structured JSON response detailing the status, actions performed, verification results, and execution duration. Minicor incorporates built-in observability features to enhance debugging and monitoring. It allows users to replay full video recordings of every automation run and provides direct failure notifications through channels like Slack. Debugging is facilitated by providing execution context, including screenshots and detailed execution logs, enabling teams to debug with context rather than guesswork. The platform features a visual workflow editor where users can define multi-step processes, such as opening an application, navigating to specific screens, filling out form fields, and verifying the submission. Execution history is maintained, allowing review of previous runs, such as monitoring the status of multiple past executions. From an architectural perspective, the system is designed to handle complex data exchange efficiently. The system allows an AI system to send a single request containing the workflow ID and necessary data, which the Desktop Service routes to the appropriate VM for agent navigation and data interaction. This streamlined process bypasses the need for external Software Development Kits or browser extensions for control. Furthermore, the entire platform maintains high security and compliance standards, being certified for SOC 2 Type II and HIPAA compliance. For on-premise deployments, the entire platform is containerized and operates strictly within the organization's network perimeter, ensuring that sensitive data does not leave the security boundary. The platform is applicable across a wide spectrum of industries and systems. Minicor supports any legacy desktop or web application, including complex healthcare systems like Epic and Cerner, dental practices systems such as Open Dental, automotive systems like CDK Global, supply chain platforms like SAP, and financial services applications. By automating interactions with these diverse systems, organizations can achieve rapid deployment, often moving from zero to production in weeks, contrasted with traditional integration approaches that commonly require four or more months. |