Automation Anywhere vs Salesforce (2026)

Agents and bots do the work inside the customer's systems. The platform combines goal-based AI agents, RPA, APIs and human expertise, connects to ERPs and CRMs through pre-built connectors and open APIs, and runs agentic AI over existing RPA and BPM systems to handle exceptions and unstructured data.

AI Agent Studio agents orchestrate across bots, APIs, documents and other agents in real time, each with assigned tool access, and the Control Room reference includes a Device API and work item queue endpoints. Automation Anywhere does not say which ERPs and CRMs are covered and gives no connector list.

Sourceautomationanywhere.com/products/agentic-process-automation-system, /products/ai-agent-studio; readread 2026-09-17

Agents act on the customer's systems through actions. Agentforce Builder connects agents to data, channels and existing Salesforce assets including Flows, Apex, MuleSoft APIs and prompt templates, and out of the box actions cover common jobs such as order and reservation management.

Sourcesalesforce.com/agentforce/agent-builderread 2026-09-18

The platform orchestrates goal-based AI agents, RPA, APIs and human expertise in one unified platform. Its goal-driven AI agents plan, execute and self-heal without manual direction, and agentic AI sits over existing RPA and BPM to handle exceptions and unstructured data that traditional bots cannot.

In AI Agent Studio, agents handle reasoning and human in the loop steps and orchestrate across bots, APIs, documents and other agents in real time, and customers place agents within end-to-end workflows using drag-and-drop tools with AI-powered next action suggestions. The Mozart Orchestrator and Automation Workspace are the named orchestration products. Agent steps are composed with deterministic bots and APIs in one workflow, and self healing covers retries. How branching and versioning work is not described.

Sourceautomationanywhere.com/products/agentic-process-automation-system, /products/ai-agent-studio; readread 2026-09-17

Agent Script defines how work moves through an agent. Every request begins at an agent router that sets initial variables and classifies the request, then the agent processes a subagent and transitions between subagents along paths the script defines. Across agents, an orchestrator agent can invoke specialized agents in other Salesforce orgs and outside Salesforce through the Agent API, and a flow action can invoke an agent as one step of a longer automation.

Sourcedeveloper.salesforce.com/docs/ai/agentforce/guide/ascript-flow.mdread 2026-09-18

Event triggers can be added and linked to a Bot Runner, and a Trigger API can also create and delete them, with trigger types including email, file and window triggers. Control Room Bot Scheduler API endpoints (/v2/schedule/automations) and an Automations API create, update, delete and return details on scheduled automations. Through Automation Co-Pilot, people can also start multi step agentic workflows in natural language inside their existing apps.

Sourcedocs.automationanywhere.com event triggers, Trigger API, Control Room APIs (index copies); automationanywhere.com/products/agentic-process-automation-system; readread 2026-09-17

Headless agents are called without a human trigger from a user interface. An Apex trigger or a scheduled job calls the agent as part of a background process, and a standard AI Agent action lets any flow, including one started by a record change, invoke an active agent with a prompt.

Sourcedeveloper.salesforce.com/blogs/2025/04/build-headless-agents-with-the-agent-api.htmlread 2026-09-18

AI agents are built on a foundational model chosen for the use case and grounded in the customer's company data, which customers connect alongside their choice of LLM. The platform also ships Document Automation and Process Discovery products that work over customer content. Automation Anywhere describes no index, graph or embeddings layer that persists and stays queryable over the customer's knowledge.

Sourceautomationanywhere.com/products/ai-agent-studio; readread 2026-09-17

Agents are grounded in the customer's own material. An Agentforce Data Library connects an agent to trusted unstructured or semi structured sources, such as files or a knowledge base, which the customer adds to the library and the agent retrieves from when it answers, and new material is added to the library with no retraining.

Sourcedeveloper.salesforce.com/docs/ai/agentforce/guide/headless-examples-agent-adl.mdread 2026-09-18

Public pages for AI Agent Studio, the platform and the Trust Center describe prompts, tools, roles and templates but no agent memory, session state or learned context, and the product documentation does not confirm whether agents keep memory.

Sourcedocs.automationanywhere.com (client-rendered); automationanywhere.com/products/ai-agent-studio; triedread 2026-09-17

Agent Script variables carry context through a session, shared by every subagent, and a session ID carries it across invocations.

Beyond the session, Data 360's Agentic Memory extracts a summary of each conversation, profile facts and preferences from a registered agent's conversations. Memory is scoped to a data space and a unified individual, and agent level access control decides which agents read which memories. Conversation summaries and profile facts persist until deleted, while preferences fade unless they are reinforced, and an expired entry can be recovered by updating its validTo field.

Memory Management APIs create, retrieve, update and delete memory entries, which is how a product lets people view, edit and delete their own facts and preferences. Deletion is permanent.

Sourcedeveloper.salesforce.com/docs/data/agentic-memory-context memory overview and memory managementread 2026-10-25

In AI Agent Studio, goal driven agents powered by the Process Reasoning Engine handle reasoning and human-in-the-loop steps and orchestrate across bots, APIs, documents and other agents in real time. Customers plug AI agents into workflows with 100% oversight, and built-in governance and guardrails are meant to keep AI actions clear, safe and accountable. The platform orchestrates human expertise alongside agents and bots in one workflow, and AI Governance is a named product. Where a checkpoint is configured, and whether autonomy varies by risk, is not described.

Sourceautomationanywhere.com/products/ai-agent-studio, /products/agentic-process-automation-system; readread 2026-09-17

Escalation to a person is a built in agent utility. The utils.escalate call hands the conversation to a human service rep through an Omni-Channel routing connection the customer defines, and instructions tell the agent when to call it. Agent Script structures what the agent may do, with every request passing an agent router that classifies it before any subagent runs and explicit transitions between subagents, and Salesforce's own guidance recommends a person check model responses before they are shared in most use cases.

Sourcedeveloper.salesforce.com/docs/ai/agentforce/guide/ascript-ref-utils.mdread 2026-09-18

Automation Anywhere holds SOC 2 Type II, ISO 27001 and GDPR certifications, and its Trust Center lists ISO 27001, SOC, HIPAA compliance, HITRUST certification, ISO 9001, ISO 42001, ISO 22301 and Cyber Essentials, with vendor penetration test, Veracode and Black Duck reports available on request.

Access is protected by RBAC (role based access control), end to end encryption and comprehensive audit trails, with security embedded at the infrastructure, data and agent levels. AI Agent Studio gives each agent a specific role, goal and tool access and limits models to an approved list, so access is controlled at user, tool and model level. The Trust Center lists SOC without a type, and the platform page supplies Type II.

Sourceautomationanywhere.com/products/agentic-process-automation-system, /trust-center; readread 2026-09-17

Salesforce holds attestations scoped to the agent layer itself, including a SOC 2 report and a SOC 3 report for the Einstein Platform and Agentforce on Hyperforce, an ISO 42001 certificate covering the Salesforce AI Platform and Agentforce, a PCI attestation of compliance for Data Cloud, the Einstein Platform and Agentforce, a C5 report, an IRAP assessment and Spain ENS High for the Einstein Platform and Agentforce, alongside FedRAMP High and Moderate and ISO 27001 categories.

Sourcecompliance.salesforce.com/enread 2026-09-18

AI Agent Studio lets customers monitor every move agents make, includes Monitor and audit as a step in how it works, and shows what goes in and comes out of the AI models, with built in governance and guardrails meant to keep AI actions clear, safe and accountable. Security includes comprehensive audit trails at the agent level, and the Control Room documentation includes an Audit API to retrieve audit log data, plus audit event and export topics. Automation Anywhere describes this monitoring only in marketing terms and does not show run level detail.

Sourceautomationanywhere.com/products/ai-agent-studio, /products/agentic-process-automation-system; docs.automationanywhere.com Audit API (index copy); readread 2026-09-17

Every step of an agent interaction is recorded. The Agentforce Session Tracing data model and the Generative AI Audit and Feedback data model capture turns, messages, LLM calls, actions, metric scores and feedback, and the Session Trace OTel API exports a full session as OpenTelemetry JSON for any observability tool or OTLP collector. Trace files can also be managed from the command line during development, and the Einstein Trust Layer keeps an audit trail of generations.

Sourcedeveloper.salesforce.com/docs/ai/agentforce/guide/otel-api.mdread 2026-09-18

The platform is sold as a cloud service, through a Cloud Service product and a free Cloud Community Edition, with no published hosting region, residency statement or option to run in the customer's environment. A Deployment Models topic exists in the documentation, but cloud regions and private cloud or on premises options for Automation 360 are not confirmed.

Sourcedocs.automationanywhere.com Deployment models (client-rendered); automationanywhere.com home and product pages; triedread 2026-09-17

Salesforce runs on Hyperforce, its public cloud infrastructure, where customer data is generally stored in the country where the customer's org is located, provided the services in use are available in that Hyperforce country, and a secure backup of an org can be held in a secondary Hyperforce region.

Sourcesalesforce.com/platform/public-cloud-infrastructureread 2026-09-18

Solutions are offered by industry, covering financial services, healthcare, life sciences and manufacturing, and by function, covering customer service, finance and accounting, human resources, IT and service operations, and each is adopted as a separate engagement.

AI Agent Studio includes a Templatize build step that saves prompts as templates for future use, agents can be customized from templates or the customer's own prompts, and prompts and agents are shared across automation teams.

The Aisera branded solutions in the menu come from the acquired company, and whether the solutions or the Bot Store include adoptable prebuilt agents is not described.

Sourceautomationanywhere.com home Solutions menu, /products/ai-agent-studio; readread 2026-09-17

Agentforce Builder starts agents from out of the box subagents, actions and templates that cover common jobs such as order management, reservation management and general FAQs, which the customer then extends and refines with natural language instructions.

Sourcesalesforce.com/agentforce/agent-builderread 2026-09-18

In AI Agent Studio, the customer chooses an AI model for each agent from an approved list, connects it to company data and can test different models to see which fits best. Custom AI models and integrations with third-party AI platforms are supported, and the Control Room API reference includes an AI Agent Studio API for defining and connecting custom model connections. The providers are not named.

Sourceautomationanywhere.com/products/ai-agent-studio; docs.automationanywhere.com Control Room APIs (index copy); readread 2026-09-17

An administrator sets the org level model for Agentforce on the Agentforce Agents page in Setup, and Agent Script overrides it for a specific agent or subagent with a model configuration naming any supported model. The supported list spans Salesforce managed models from OpenAI, Anthropic on Amazon Bedrock, Google Gemini on Vertex AI and Amazon Nova, most running inside the Salesforce trust boundary, and customers can also bring their own model.

Sourcedeveloper.salesforce.com/docs/ai/agentforce/guide/ascript-model.mdread 2026-09-18

The versioned Control Room API reference lists endpoint modules including an Authentication API, Audit API, Device API, Automations (Bot Scheduler) API, Trigger API, Policy Management API, an AI Agent Studio API for defining and connecting custom model connections, and an API Task Execution API that generates an execution URL and token to start an API-task execution through an API endpoint. The platform integrates through pre-built connectors and open APIs, and a free Community Edition is offered to developers. No MCP support is described, and a Package SDK appears in release notes.

Sourcedocs.automationanywhere.com Control Room APIs (index copy); automationanywhere.com/products/agentic-process-automation-system; readread 2026-09-17

The Agent API opens a conversation with an agent, sends prompts synchronously or as a stream of server sent events marking progress, text chunks and the end of each turn, and closes the session, authenticated through a connected app with OAuth scopes, with a public Postman collection and Node.js client. Inside the platform, an invocable action runs any active agent from Apex or Flow, and Salesforce Hosted MCP Servers expose Salesforce data and automation to external AI agents and LLM applications over MCP.

Sourcedeveloper.salesforce.com/blogs/2025/04/build-headless-agents-with-the-agent-api.htmlread 2026-09-18

AI Agent Studio's build steps are Select and connect, Tune and test, Templatize, Plug in to workflows, and Monitor and audit. In Tune and test, builders adjust prompts to meet their exact requirements and test for accuracy, and in Select and connect they test different models to see which is the best fit, all before an agent goes into a workflow. The AI Governance product sits alongside. Whether results are scored and kept is not described.

Sourceautomationanywhere.com/products/ai-agent-studio; readread 2026-09-17

Agents are tested before release with scored test cases. A test spec sets out test cases as utterances with the expected agent behavior, can set context variables and include ready made metrics in the output, and is created and run in a development org from the command line or the Agentforce Testing Center, with pass and fail results.

Sourcedeveloper.salesforce.com/docs/ai/agentforce/guide/agent-dx-test.mdread 2026-09-18

Automation Anywhere controls real application interfaces through Robotic Process Automation, a product line that includes RPA Software, Automation Workspace and Automation 360. The platform combines goal-based AI agents, RPA, APIs and human expertise, with agentic AI running over existing RPA and BPM systems to handle exceptions and unstructured data that traditional bots cannot. Software robots do the screen level work inside the same process. How the robots operate interfaces, and how often they break when an interface changes, is not described.

Sourceautomationanywhere.com/products/agentic-process-automation-system, home product menu; docs.automationanywhere.com (client-rendered); readread 2026-09-17

Operating a browser or desktop interface is not part of Agentforce. Agents act through actions, flows, Apex, APIs, prompt templates and MCP tools, reached through the agent router and subagents in Agent Script.

Sourcedeveloper.salesforce.com/docs/llms-genai.txtread 2026-09-18