- Erste Schritte
- Best Practices
- Organisationsmodellierung im Orchestrator
- Beste Praktiken für die Automatisierung (Automation Best Practices)
- Optimieren von Unattended-Infrastruktur mithilfe von Maschinenvorlagen
- Organisieren von Ressourcen mit Tags
- Exportieren von Rastern im Hintergrund
- Durchsetzung der Governance der Integration Service-Verbindung auf Benutzerebene
- Mandant
- Über den Kontext „Mandant“
- Suche nach Ressourcen in einem Mandanten
- Verwaltung von Robotern
- Verbindung von Robotern mit Orchestrator
- Speicherung von Roboterzugangsdaten in CyberArk
- Speichern der Kennwörter von Unattended-Robotern im Azure Key Vault (schreibgeschützt)
- Speichern der Anmeldeinformationen von Unattended-Robotern im HashiCorp Vault (schreibgeschützt)
- Speichern der Anmeldeinformationen von Unattended-Robotern im AWS Secrets Manager (schreibgeschützt)
- Löschen von getrennten und nicht reagierenden Unattended-Sitzungen
- Roboter-Authentifizierung
- Roboter-Authentifizierung mit Client-Anmeldeinformationen
- Konfigurieren von Automatisierungsfunktionen
- Solutions (Lösungen)
- Audit
- Erzwungene Verschlüsselung
- Einstellungen
- Registrierung
- Benachrichtigungen
- Events
- Anzeigen und Zugreifen auf Benachrichtigungen
- Anzeigen und Zugreifen auf E-Mail-Benachrichtigungen
- Es werden nur ungelesene Benachrichtigungen angezeigt
- Alle Benachrichtigungen als gelesen markieren
- Alle Benachrichtigungen löschen
- Löschen von Benachrichtigungen
- Abonnieren von Ereignissen
- Abbestellen von Ereignissen
- Ordnerkontext
- Prozesse
- Jobs
- Apps
- Auslöser
- Protokolle
- Überwachung
- Indizes
- Warteschlangen
- Assets
- Über Assets
- Verwalten von Assets in Orchestrator
- Verwalten von Assets in Studio
- Speichern von Assets im Azure Key Vault (schreibgeschützt)
- Speichern von Assets im HashiCorp Vault (schreibgeschützt)
- Speichern von Assets im AWS Secrets Manager (schreibgeschützt)
- Speichern von Assets in Google Secret Manager (schreibgeschützt)
- Verbindungen
- Geschäftsregeln
- Speicher-Buckets
- Agent-Gateway
- Testverfahren in Orchestrator
- Ressourcenkatalogdienst
- Integrationen
- Fehlersuche und ‑behebung
Autopilot Chat in Orchestrator
Use Autopilot Chat in Orchestrator to build, operate, and troubleshoot automations and tenant resources through natural language.
Die Verfügbarkeit der Funktionen hängt vom Cloud-Angebot ab, das Sie verwenden. Weitere Informationen finden Sie auf der Seite Funktionsverfügbarkeit.
Autopilot Chat is an AI assistant built into Orchestrator. It answers questions from the UiPath documentation and the web, and it works directly with your Orchestrator tenant, so you can manage resources, investigate problems, and plan automations through natural language.
This page covers what is specific to Orchestrator. For the chat features shared across UiPath products, such as file attachments, model selection, and chat history, refer to the Autopilot user interface page.
What Autopilot can do in Orchestrator
Autopilot in Orchestrator can help you with the following:
- Build automations: RPA workflows, agents, API workflows, Maestro flows, BPMN processes, case plans, coded apps, coded functions, business rules, and connectors. Projects are created and edited in Studio Web.
- Operate your tenant: manage Orchestrator processes, jobs, queues, assets, triggers, machines, folders, and storage buckets, as well as Integration Service connections, Data Fabric records, and licensing.
- Work with AI and documents: create Document Understanding projects, taxonomies, extraction models, knowledge bundles, context grounding configurations, and agent integrations.
- Diagnose problems: investigate failed jobs, access issues, connector errors, workflow faults, performance problems, policy conflicts, and deployment failures, using the available runtime and trace evidence.
- Review and improve: audit workflows, agents, flows, solutions, code structure, governance, reliability, and best practices.
- Plan solutions: turn an idea, a standard operating procedure, a transcript, or a process definition document into a solution design document, an implementation plan, an automation blueprint, an effort estimate, or an automation discovery report.
- Manage lifecycle and governance: package, publish, deploy, and migrate automations, configure policies, manage permissions, inspect audit events, and support CI/CD pipelines.
- Analyze operations: use Insights and Process Mining to examine job performance, queue SLAs, failure patterns, process bottlenecks, and automation opportunities.
- Create reusable skills: turn a repeatable procedure into a reusable agent skill or workflow.
Actions that affect cloud resources or external systems may require your confirmation, depending on the approval mode you select.
Die AI Trust Layer-Richtlinie
Um die von Autopilot bereitgestellten KI-Funktionen zu nutzen, stellen Sie sicher, dass eine AI Trust Layer-Richtlinie aktiv ist und für Ihren Mandanten bereitgestellt wurde.
Erfahren Sie, wie Sie:
- Erstellen Sie eine Richtlinie
- Konfigurieren Sie die Richtlinie für „AI Trust Layer“.
- Deploy the policy to the tenant where you want to use Autopilot
Approval modes
The approval mode controls when Autopilot asks for your confirmation before it runs an operation. Select the approval mode from the dropdown below the chat input.
| Genehmigungsmodus | Verhalten |
|---|---|
| Vorsichtig | Asks before every operation. This is the safest option and shows the most approval prompts. |
| Adaptive | Auto-approves known safe operations. This is the recommended option. Turn on Smart approvals to also approve the operations that the model judges safe. This shows the fewest approval prompts, but the model can misjudge. |
| Vollständiger Zugriff | Shows no approval prompts. Use it only in trusted environments. |
Benutzeroberfläche
Autopilot in Orchestrator bietet die folgenden Optionen:
- New chat: starts a fresh conversation and saves the previous one in your chat history. Use it when you want to switch topics, so your old prompts and answers do not affect the new interaction.
- Chat history: displays and reopens past conversations with Autopilot from the last 30 days.
- Settings: controls how Autopilot behaves in Orchestrator, organized in the following tabs:
- Tools: turns the tools that Autopilot can use on or off, such as web search and web reader.
- Skills: turns the built-in skills that Autopilot can draw on on or off for the current session.
- MCP servers: connects MCP servers to expose additional tools to Autopilot.
- Personalization: adds a custom instruction to every prompt, turns follow-up suggestions on or off, and adjusts the communication style of the assistant.
- Security: lists the websites that Autopilot can use without asking for confirmation, and the websites that Autopilot won't access.
- Annotate page element: lets you select an element on the Orchestrator page and add an optional comment about it to your prompt. While annotating, the button changes to Stop annotating.
- Add (+): attaches files for Autopilot to work with, and runs commands such as
/usage, to view your usage for the current period, or/goal, to let Autopilot keep working until the goal is met. - Approval mode: sets when Autopilot asks for confirmation. For details, see Approval modes.
- Model picker: selects the model and the effort level that Autopilot uses.
Opening and closing Autopilot Chat
- Öffnen Sie in Ihrer Automation Cloud™-Organization den Orchestrator.
- In the Orchestrator header, select the Autopilot icon
. The Autopilot panel opens next to the page content. - Enter your request in the chat input, or select one of the suggested prompts.
To close the panel, select the Autopilot icon again.