IBM watsonx Orchestrate vs UiPath (2026)

Tools are implemented and called by agents to invoke external APIs and services, a tool node places a tool inside an agentic workflow, and the ADK covers tool definition, connections to external systems with their own credential and OAuth token handling, an OpenAPI plugin and MCP based workflows.

Sourceibm.com/products/watsonx-orchestrate/developersread 2026-09-17

Tools connect an agent's reasoning to real actions, giving it access to data in business applications and the ability to execute functions, and an agent can only act through tools it has been explicitly granted, described in the system prompt with when to use them. Integration Service adds an MCP Connector that lets workflows call tools from any compatible MCP server as a native activity, Database Hub to read and write systems such as SQL Server, Oracle and Databricks without custom code, and Relay to reach on premises systems from cloud automations.

Sourcedocs.uipath.com/agents (agent tools) and uipath.com/newsroom/uipath-platform-updates-fusionread 2026-10-24

Agentic workflows are assembled from node types, including an agent node that hands a step to another agent, branch, parallel, foreach and decision nodes, a script node, a timer node that sets a delay between actions, and a user activity node. Builders work with the Agent Development Kit for pro code control, Langflow for visual design and Flow Builder for rapid orchestration, with branching logic and long running flows.

Sourceibm.com/products/watsonx-orchestrate/developersread 2026-09-17

UiPath orchestrates agents of every origin, UiPath system agents, low-code agents, coded agents and external agents, inside one platform, where agents work alongside activities, other agents, apps, human tasks and Maestro for business process orchestration, and are managed, scheduled and monitored through Orchestrator.

Sourcedocs.uipath.com/agents/automation-cloud/latest/user-guide/agent-capabilities-in-the-uipath-platform.mdread 2026-09-18

Agentic workflows and agents can be scheduled to run automatically at specific times or intervals once schedulable is set on the workflow or agent, with the platform handling cron patterns, timezones and execution limits and the schedule running without manual intervention, and separately each channel a buyer configures (web chat, Twilio WhatsApp, Twilio SMS, Slack, Genesys Bot Connector, Facebook Messenger, Microsoft Teams) provides its own event URL to receive messages.

Sourcedeveloper.watson-orchestrate.ibm.com/scheduling/overview.mdread 2026-09-17

Agents deploy as processes in Orchestrator, which schedules and launches jobs through time triggers at regular intervals and queue triggers that start the automation whenever new items are added to a queue, so agents run without anyone asking.

Sourcedocs.uipath.com/orchestrator/automation-cloud/latest/user-guide/about-triggers.mdread 2026-09-18

Agents use knowledge bases to search and return grounded content. Knowledge bases are defined in YAML or JSON and imported through the ADK CLI, in two forms: built-in, backed by a Milvus instance and populated with documents the customer imports, or external, pointing at the customer's own Milvus or Elasticsearch instance.

Sourcedeveloper.watson-orchestrate.ibm.com/knowledge_base/overview.mdread 2026-09-17

An index added as agent context gives the agent access to a permissioned knowledge base of the customer's own content, and the builder configures the search strategy and scope that decide what the agent retrieves.

Sourcedocs.uipath.com/agents/automation-cloud/latest/user-guide/agent-contexts.mdread 2026-09-18

The Agentic SDK ships memory APIs that persist user information and retrieve it during later agent execution, against /memories and /memories/search endpoints: add_messages writes, search, list and retrieve read, and delete and delete_all remove, with memories typed (for example preference), tagged with metadata, scoped by agent and run, and carrying a sensitivity classification.

Sourcedeveloper.watson-orchestrate.ibm.com/sdk/memory.mdread 2026-09-17

Agent memory, in preview, keeps memory spaces that several agents can share. Episodic memory holds the inputs, outputs and outcomes of past cases with the feedback attached to them, and items enter it when a person promotes feedback from an agent run. Escalation memory stores how a person resolved an escalation in Action Center.

Before each run a recall step embeds the current inputs, finds similar items and adds them to the agent's context as examples, and before raising an escalation the agent checks for a matching resolution. Memory items can be inspected, edited and deleted from the Memory spaces tab. Nothing is written to memory by the agent on its own, and no retention period or expiry is stated.

Sourcedocs.uipath.com agents user guide, agent memoryread 2026-10-25

Agentic workflows carry a user activity node, an interactive step where the workflow waits for a person to provide or receive information, either one field at a time in conversation or through a form that collects several fields at once, with node owners, input and output schemas, and fields that select approvers from a user list. Workflows pause for human validation, capture context and resume automatically, with traceability ready for audit.

Sourcedeveloper.watson-orchestrate.ibm.com/tools/flows/user_activities_node.mdread 2026-09-17

Escalations, powered by Action apps in Action Center, let a developer have the agent validate business details with a person before running a tool, request approval, or ask for manual assistance on a business exception. Each escalation carries a prompt that tells the agent when to raise it, and is routed to a specific user, to the least-loaded member of a group, to all group members, or round robin. Runtime Checker validates agent behavior against defined policies while the agent is running, not only when it was built, and an LLM as judge guardrail evaluates outputs against custom criteria.

Sourcedocs.uipath.com/agents (agent escalations) and uipath.com/newsroom/uipath-platform-updates-fusionread 2026-10-24

Security and access control is named as a feature from the entry plan, and the Premium plan adds data isolation, a dedicated data space, end-to-end protection and a HIPAA-ready option. IBM does not say which identity providers and provisioning standards are supported, or whether access control reaches users, teams, tools, knowledge sources and model usage, and it publishes no certification or audit report scoped to the product.

Sourceibm.com/products/watsonx-orchestrate/pricingread 2026-09-17

UiPath's trust portal publishes a SOC 2 Type 2 report, an ISO 27001 certificate, a HITRUST certificate, C5, IRAP, HIPAA and Spain's ENS certificate of conformity. The customer-facing controls on the agent layer are Automation Ops policies applied through the AI Trust Layer at tenant level, which control whether users can access and use agents across the organization.

Sourcetrust.uipath.comread 2026-09-18

Traces visualize the complete path a request takes through an agent, with each observation recording a distinct unit of work, and they are reachable four ways: the orchestrate observability traces CLI command, the ADK Python library, the product interface and the API.

Sourcedeveloper.watson-orchestrate.ibm.com/traces/overview.mdread 2026-09-17

Every agent run leaves a trace, a detailed record of the steps taken, data processed, decisions made and results generated, with timestamps, errors, inputs and outputs, so a failed tool call can be inspected at the exact step with its inputs, outputs and error. UiPath positions traces for debugging, performance analysis of latency, errors and throughput, and compliance and auditing as a verifiable record of what the agent did, when and how.

Sourcedocs.uipath.com/agents/automation-cloud/latest/user-guide/agent-traces.mdread 2026-09-18

The product runs as a fully managed service on IBM Cloud or on AWS, or in the customer's own on premises environment, and the ADK supports on premises setups and air gapped environments.

Sourceibm.com/products/watsonx-orchestrate/pricingread 2026-09-17

UiPath comes in two delivery models, Automation Cloud, its cloud service, and Automation Suite, which runs the platform on the customer's own premises, each with its own documentation set. Automation Suite now runs on Linux with the full agentic AI stack, so agents, robots and people work inside infrastructure the enterprise already operates.

Sourcedocs.uipath.com/llms.txt and uipath.com/newsroom/uipath-platform-updates-fusionread 2026-10-24

An agents and tools catalog is a named feature of every plan, with ready-to-use domain agents added at the Standard plan, and the product API exposes catalog management operations for listing the applications and skills available from it. Partner built agents are sold beside it as separately metered add ons, among them a web research agent, a sales intelligence agent, a messaging agent and a document question answering agent.

Sourceibm.com/products/watsonx-orchestrate/pricingread 2026-09-17

UiPath ships agent templates, available on the Agents page in Automation Cloud and in UiPath Marketplace, each preconfigured for a specific task with a system prompt, activity tools with descriptions and starter evaluation sets, needing only minimum configuration to run and open to modification.

Sourcedocs.uipath.com/agents/automation-cloud/latest/user-guide/using-agent-templates.mdread 2026-09-18

Models from multiple providers are imported as virtual models and assigned to an individual agent, and model policies coordinate several of them with a load-balance, fallback or single strategy, weights per target, and retry on named HTTP status codes; a policy is assigned to an agent the same way a single model is. On the product side, the same surface is an AI Gateway for integrating models from several providers with configurable routing policies. Model policies are in public preview, while assigning a model to each agent is not in preview.

Sourcedeveloper.watson-orchestrate.ibm.com/llm/model_policies.mdread 2026-09-17

The builder chooses the model for each agent at design time and refines that choice once evaluations show how different models perform on accuracy, stability, speed and cost, and a custom model can be configured for Context Grounding through the admin guide's LLM configuration.

Sourcedocs.uipath.com/agents/automation-cloud/latest/user-guide/choosing-the-best-model-for-your-agent.mdread 2026-09-18

IBM publishes an Agent Development Kit with a Python SDK and an orchestrate command line (agents import, evaluations, observability traces, models policy), a machine readable documentation index, and an OpenAPI specification for the server API with reference pages for each operation on agents, flows, connections, catalog management, evaluation and monitoring.

Sourcedeveloper.watson-orchestrate.ibm.com/llms.txtread 2026-09-17

With a UiPath SDK, developers build coded agents in their own IDE, package them and deploy them to Orchestrator as standard packages, where they run as processes under the same governance as other automations, and they can connect the project to Studio Web for debugging, evaluation and publishing.

Sourcedocs.uipath.com/agents/automation-cloud/latest/user-guide/about-coded-agents.mdread 2026-09-18

The ADK ships an evaluation framework that assesses agent behavior by simulating user interactions and comparing the resulting trajectories step by step against reference data the customer generates, records or writes by hand, logging every deviation from the expected path and returning a metrics summary saved as CSV. The cycle before deployment is develop, evaluate, analyze, improve, with an analyze command for reading results and a rubric, and evaluations run against the local Developer Edition and against draft environments of SaaS and on premises instances.

Sourcedeveloper.watson-orchestrate.ibm.com/evaluate/overview.mdread 2026-09-17

Evaluations check agents against defined cases before and during release. An evaluation pairs an input with an evaluator that checks the agent's output or trajectory against the expected result, evaluation sets group evaluations, and each run produces a score from 0 to 100 with traces of the evaluation runs to diagnose failures and re-run. Evaluations can be created from correct past runs, and evaluation results guide model choice.

Sourcedocs.uipath.com/agents/automation-cloud/latest/user-guide/agent-evaluations.mdread 2026-09-18

Agents act through tools that call APIs and services, connections to enterprise systems, MCP and the channels a buyer configures. IBM does not describe browser, desktop or remote computer control.

Sourceibm.com/products/watsonx-orchestrate/pricingread 2026-09-17

Robots operate real application interfaces in attended and unattended automation, and Healing Agent autonomously detects and recovers from UI automation failures caused by changing interfaces, network throughput and evolving UI frameworks, using just-in-time analysis of the automation to choose a recovery strategy.

Sourcedocs.uipath.com/agents/automation-cloud/latest/user-guide-ha/what-is-healing-agent.mdread 2026-09-18