Qualtrics vs Salesforce (2026)

Once approved, Experience Agents perform ticket actions, including status updates, escalation and reassignment, and they create tickets of their own. These are writes into the XM Platform's system of record, not lookups. They read across survey, call, chat and digital experience analytics data, so a single agent draws on several data types in one decision.

Outside the agent layer, the platform has a public API with SDKs, Postman collections and workflow tooling for managing surveys, users and results, which gives a customer's own systems a route in and out. The agent pages show no connectors into third party systems of record such as a CRM or commerce platform, so the agents act in Qualtrics' own platform, not the customer's wider stack.

Sourcequaltrics.com/support/experience-agents/experience-agents-overviewread 2026-09-05

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

In Agent Studio a specific agent is created for each workflow and trained on that company's policies and runbooks. The Ticketing Agent runs a real sequence with conditions. It summarizes the ticket, recommends an action, drafts the response, waits for approval, then executes the approved ticket action, escalating or reassigning where the case requires it.

Agents coordinate handoffs to people when oversight is needed. Survey agents resolve an issue inside the response flow before it becomes a ticket at all, so orchestration spans two surfaces. Experience Agents are built on LangChain's LangGraph platform, a framework under the hood that the customer does not work with directly.

Sourcequaltrics.com/support/experience-agents/experience-agents-overviewread 2026-09-05

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

Experience Agents act inside a survey the moment feedback is given, so the trigger is the customer's own words arriving, not a ticket being raised. Digital agents detect friction across a journey and step in instantly, so the platform starts a conversation off a behavioral signal the customer never reported. Ticketing agents fire within agent-enabled ticketing workflows as cases arrive and change state. All three starts are automatic and event-driven. Channels are narrow, running through surveys, tickets and instrumented digital touchpoints, not voice, SMS or messaging apps such as WhatsApp.

Sourcequaltrics.com/platform/experience-agentsread 2026-09-05

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

Qualtrics grounds Experience Agents in its proprietary human experience data, roughly twenty years of customer and employee feedback captured across hundreds of thousands of touchpoints. This maintained corpus is the product's main differentiator.

Qualtrics sets the agents against transactional agents built on systems of record, claiming the agent understands why an issue occurs, not only what to execute. Each agent is also trained on the individual company's policies and runbooks, so answers reflect that business's rules, not generic guidance. Agents draw on survey, call, chat and digital experience analytics across the platform. The corpus and the runbooks persist and are maintained.

Sourcequaltrics.com/platform/experience-agentsread 2026-09-05

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

Context persists in the platform, but the agent has no memory layer of its own. Experience Agents read an individual's prior interactions from the XM Platform and predict trends from that experience context, so an agent meeting a customer for the second time knows what happened the first, and the record survives across sessions because the platform holds it.

That is retrieval from a store the customer already owns, not state the agent forms and keeps. There is no memory product, scope, lifetime or deletion control for state the agent holds. Qualtrics says each interaction adds to the intelligence, which describes learning absorbed into a model, not a memory layer.

Sourcequaltrics.com/platform/experience-agentsread 2026-09-05

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

A person approves before the agent sends or acts. The Ticketing Agent recaps the situation, recommends an action and drafts the email response, and a person approves before anything is sent or executed. Only once approved does the agent carry out ticket actions such as status updates, escalation or reassignment. Compensation has limits.

Agents apply make goods within guardrails the brand sets, so a service-recovery gesture cannot exceed what the business authorized. Handoffs are coordinated when human oversight is required, and Qualtrics cites Stanford Health Care testing patient care coordination explicitly under human oversight.

Sourcequaltrics.com/support/experience-agents/experience-agents-overviewread 2026-09-05

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

Attestations include SOC 2 Type II covering security, availability and confidentiality, ISO/IEC 27001, 27017, 27018 and 27701, and ISO/IEC 42001:2023, the first international standard for AI management systems.

Qualtrics also holds HITRUST certification for healthcare, TISAX for the German automotive sector (retrievable through the ENX portal), and PCI DSS certification on the XM Discover voice-of-customer integration for call recordings containing card data. Customers can get the reports through a Trust Center and a self-service repository for the SOC 2 report, the ISO certificates, the HITRUST validated report and the PCI attestation.

Access controls include single sign-on authentication, enforced multi-factor authentication, user log and account activity monitoring through the Control Panel and APIs, password complexity controls, PII redaction and collection restriction, and customer-controlled retention with one-click deletion of personal data across responses, tickets and contacts for GDPR erasure requests. API-side controls add IP allowlisting and mutual TLS. FedRAMP High and IL4 authorizations add sovereign delivery environments.

Sourcequaltrics.com/security-statementread 2026-09-05

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

Every action an Experience Agent performs on a ticket is visible in the AI activity log within Tickets, and the agent's recommendations and actions appear directly inside the tickets it creates, so a supervisor can reconstruct what the agent did and why without leaving the workflow. Underneath, the platform offers platform-level user log and account activity monitoring through the Control Panel and APIs. That log does not capture all read access to brand directories, and Qualtrics routes such requests to its Security Operations Center.

Sourcequaltrics.com/support/experience-agents/experience-agents-overviewread 2026-09-05

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

Qualtrics has a dedicated Data Sovereignty page. It operates multiple data centers with load balancing between them and independent third-party audits of the data center providers, and an additional isolation tier keeps a customer's data fully walled off in a virtual environment. Placement is set per account and visible to the customer.

Each account sits in a specific named data center, and developers select the API server that matches it. Distinct government environments sit above that, FedRAMP High for federal civilian agencies and Impact Level 4 for the Department of War, so a public-sector customer is deployed into a separately authorized environment, not the commercial one.

Sourcequaltrics.com/platform/security/data-sovereigntyread 2026-09-05

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

Specialized agent types ship ready to configure, but they are types, not a catalog. Survey Experience Agents act inside a response the moment feedback is given, and a Ticketing Agent summarizes, recommends actions, drafts responses and performs approved ticket actions, with more specialized agents interacting across touchpoints.

A customer starts from a working agent shape, not an empty canvas. There is no template gallery, agent library or marketplace that a customer browses and puts into service, and the build route points the other way, since a specific agent is created for each workflow in Agent Studio.

Sourcequaltrics.com/support/experience-agents/experience-agents-overviewread 2026-09-05

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

There is no model selection, routing configuration or bring-your-own-key path, and no model provider is named. Experience Agents are built on LangChain's LangGraph platform, an orchestration framework Qualtrics builds on, not a model the customer can choose or point their own account at. The open source framework Qualtrics says it is developing concerns interoperability between agents, not choice of model, and it has not shipped.

Sourcequaltrics.com/platform/experience-agentsread 2026-09-05

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

Qualtrics runs a Developer Portal at developer.qualtrics.com with registration, tooling and documentation for building in its ecosystem. A public API reference at api.qualtrics.com offers development kits for building applications, public API collections in Postman, and advanced connection controls including IP allowlisting and mutual TLS.

The support documentation has an interactive API tester with per-request parameter tables, cURL examples and code snippets in Python and Java for managing surveys and users and retrieving results. All of it is reachable without a login. Qualtrics also says it is developing an open source framework for cross vendor agent interoperability, which has not shipped.

The developer surface belongs to the XM Platform, and there is no agent-specific API for authoring or invoking Experience Agents from outside, so an integrator reaches the platform, not the agents.

Sourceapi.qualtrics.comread 2026-09-05

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

There is no simulation, test environment, scenario suite, regression run or scored pre-release evaluation of an Experience Agent, and Agent Studio is an authoring environment with no published test surface. Synthetic Research, which tests CX scenarios artificially and gives feedback on them, is a preview, not a shipped capability. Qualtrics has deep testing tooling for surveys, but what it tests is a questionnaire, not the agent.

Sourcequaltrics.com/support/experience-agents/experience-agents-overviewread 2026-09-05

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

There is no browser control, hosted session or interface automation. Experience Agents act inside surveys, tickets and instrumented digital touchpoints, and their outward reach is programmatic, through approved ticket actions written into the XM Platform and the platform's public API for anything beyond it. Digital experience monitoring observes a customer's journey to detect friction, which instruments the customer's own session and does not mean the agent operates a browser.

Sourcequaltrics.com/platform/experience-agentsread 2026-09-05

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