Moveworks

Hundreds of named enterprise systems are connected, and the Assistant takes action in them, from resetting access to updating CRM records. The AI Assistant page says Moveworks "unifies hundreds of enterprise systems under one intelligent layer," and its integrations strip names systems across seven classes. Service desk and ITSM tools include ServiceNow, Jira, Freshservice, Ivanti Service Desk, BMC Helix Digital Workplace, ManageEngine and Cherwell. HR and finance cover Workday, Coupa, SAP Ariba, SAP Fieldglass and Oracle, and Salesforce covers CRM.

Identity and access providers include Ping Identity, Duo and BeyondTrust. Content and collaboration tools are Box, Confluence, Quip, Google Drive, LumApps, Igloo, Interact, Axero and Brightspot. Mail, calendar and chat cover Microsoft Outlook, Exchange Online, Gmail, Google Calendar, Google Chat and 8x8, and knowledge and data sources include Databricks, Upland RightAnswers, ComAround and BrainStorm QuickHelp, with DocuSign for signatures.

The actions are documented, not just the catalog. The Assistant is "linked to thousands of action plugins that turn intent into results" and resolves requests end to end: it can "resolve issues, reset access, and provision tools automatically" for IT, "update CRM data" for sales, and "automate expenses, purchase requests, and approvals" for finance. Those are authenticated writes into systems Moveworks does not own. The Reasoning Engine "selects the right tools" to reach the employee's goal, so each tool call is chosen at run time rather than wired in advance.

Sourcemoveworks.com/us/en/platform/ai-assistant and /platform/integrations; moveworks.com/us/en/platform/integrationsread 2026-09-12

Each request is planned at run time by the Reasoning Engine, which breaks it into steps and picks the tools to carry them out across enterprise systems. The runtime has its own product page and is described as an agentic AI system that "orchestrates enterprise-ready LLMs and built-in tools for AI agents."

Planning is an explicit step: it "identifies overall employee goals and systematically determines the steps needed to achieve them, breaking down intricate issues into smaller components," then "selects the right tools to achieve every employee's end goal," automating processes with many steps under strict security protocols and real time governance.

The platform page calls it automation across the enterprise, "planning, executing, and adapting on the fly to complete any request," and the Assistant "goes in-depth on complex asks" with extended thinking. Plans are built at run time, not coded in advance.

The participants are distinct and named. An employee asks, the Reasoning Engine plans, and agents built or installed through Agent Studio and the Marketplace carry out the work; admins govern from a central console, and plugins act in the systems of record. Customers can extend the sequence, since developers build and deploy agents that plan, reason and execute securely across enterprise systems, and a worked example covers IT, HR and finance in one conversational request. Orchestration across departments through ServiceNow is part of the acquirer's platform.

Sourcemoveworks.com/us/en/platform/reasoning-engine, /platform and /platform/ai-assistantread 2026-09-12

Answers come from the customer's own connected systems, filtered to what each asker may see, and a dedicated authoring product keeps that corpus up to date. Enterprise Search "delivers instant, context-aware answers across business apps, tailored to role, region, language, and access level without relying on exact keywords or rigid filters," over the systems the customer connects.

The AI Assistant is described as "connected to every system and content source that matters." The trust page names the mechanism: "a grounded knowledge system that allows us to ground our solutions with the most up-to-date and trustworthy documentation available," with fact verification among its AI safeguards.

The corpus persists and is maintained. Knowledge Studio exists to "generate grounded knowledge articles to fill in content gaps and improve self-service for FAQs," so the content grows over time, with gaps found from what employees actually ask. Retrieval respects permissions: answers are filtered to the asking employee's access level, so two people asking the same question can get different results. It is a retrieval structure over the customer's own estate that lasts between runs and scales far past any context window.

Sourcemoveworks.com/us/en/platform/enterprise-search, /platform/ai-assistant, /platform/ai-knowledge-management and /platform/securityread 2026-09-12

A stack of guardrails surrounds what the agents do, but no human approval is documented before they act. The trust page lists "a content moderation system to filter harmful or inappropriate content," "prompt protection to safeguard against prompt injection attacks" and "query risk assessment to identify and mitigate potential risks associated with user queries."

It adds fact verification and "identity validation that relies on deterministic systems and source-of-truth methodologies for authentication and authorization, without the use of LLMs." The security blog says the "Reasoning Engine operates within scoped action boundaries enforced at the infrastructure level." Each of these limits what the agent may do.

What is missing is a person in the loop. No approval queue, checkpoint or escalation on a risk threshold appears, and no human role is named. Nothing documents anyone signing off before the Assistant resets access, provisions a tool, updates CRM data or completes a purchase request, all actions it is advertised as taking end to end. Enterprise permissions and scoped access govern who can reach what, and governance added by ServiceNow belongs to the acquirer. The help documentation or the Agent Studio surface may describe an approval step.

Sourcemoveworks.com/us/en/platform/security and /resources/blog/why-autonomous-ai-agents-need-new-enterprise-security-modelread 2026-09-12

Ten certifications are published with their scopes, including ISO/IEC 42001 for AI management and FedRAMP, alongside named access controls and an open bug bounty. The trust page shows each with auditor marks and a line of explanation. The ISO/IEC set covers 42001, 27001, 27017, 27018 and 27701; SOC 2 Type 2 spans security, confidentiality, availability and privacy; CSA STAR Level 2 is held at Gold; and FedRAMP, GDPR and CCPA complete the list.

The controls cover access to how user data is processed and stored, encryption at rest and in transit, and masking of sensitive PII, plus "identity validation that relies on deterministic systems and source-of-truth methodologies for authentication and authorization, ensuring secure access to systems and data without the use of LLMs." The AI Assistant page adds a zero trust design on "never trust, always verify," with least privilege access and responses scoped to each person's permissions.

Assurance work is published too: application security reviews, vulnerability scanning and external penetration testing, red team exercises against enterprise and AI systems, and a HackerOne bug bounty open to all researchers at every severity level. The certification reports sit behind a Whistic profile reached through a demo request or an account executive, so the page carries the names and scopes but not the reports. Security that ServiceNow's trust and controls add belongs to the acquirer.

Sourcemoveworks.com/us/en/platform/security and /platform/ai-assistantread 2026-09-12

Every system access and data event is logged, audit trails record what the agent did, and the Assistant shows its reasoning on each request. Under Auditable transparency, "every system access and data event is logged for full visibility, and can be reviewed to support compliance and incident response."

The security blog ties this to the agent: the Reasoning Engine works within scoped action boundaries enforced at the infrastructure level, "with detailed audit trails for compliance." That is a retained, reviewable record of the agent's actions.

Reasoning can be inspected per interaction, not only in aggregate: the AI Assistant "reasons for you and shows its work," so a reader can follow how a request became an action.

Owners get a named product on top. Employee Experience Insights gives program owners tools to "identify issues, track metrics, and enhance your assistant's efficacy with robust governance and analytics," and a self service admin console centralizes configuration and monitoring. Analytics that ServiceNow adds belong to the acquirer. The Employee Experience Insights page may show what the trail exposes to an operator, such as export, retention or attribution per action.

Sourcemoveworks.com/us/en/platform/ai-assistant and /resources/blog/why-autonomous-ai-agents-need-new-enterprise-security-modelread 2026-09-12

A help page lists the Assistant's four kinds of memory, and none is a lasting store per user; conversation context is cleared every 24 hours. The Memory Constructs page says "the Moveworks AI Assistant leverages four different kinds of memory" and defines each. Semantic memory is "knowledge of the content, entities and terminology used by the organization," which is the organization's corpus and vocabulary, shared across the organization rather than kept per user. Episodic memory sounds decisive and is not.

Its summary line offers "to recall the past conversations with the user," but the section defines it as "awareness of the current context in the conversation with the user, i.e. what questions and answers have been exchanged and what decisions have been made already," which is the current conversation, not past ones. The Conversational Context page settles it: "context is automatically cleared every 24 hours from our agentic AI reasoner," so anything older than a day is gone.

Procedure memory is "knowledge of the tasks that can be performed in the environment and what business processes or rules should be followed," meaning the plugin catalog and the customer's configured rules, which the product applies.

Working memory is "awareness of what operations and processes are in progress, and which stage of completion they are at," with a framework that tracks variables such as business object IDs, which is run state inside a single execution. The runtime description agrees: the context window is "assembled fresh on each turn" and the engine "reads everything, every turn, every time." Nothing is carried in from an earlier store, and that absence shows on a page written to list the platform's memory types.

Sourcehelp.moveworks.com/agent-studio/agentic-ai/memory-constructs, /assistant-behavior-framework and /agentic-reasoning-engine/how-the-reasoning-engine-worksread 2026-09-13

Hosting is on AWS across seven named regions, with a commitment that customer data never leaves its region and a dedicated GovCloud environment for federal clients. The trust page names three regions in North America (US Commercial, US GovCloud and Canada), EU Commercial and the UK in Europe, and Australia and Japan in Asia Pacific. The residency commitment is explicit and unconditional: "All customer data remains in its region and is never transferred elsewhere." It is stated as a property of the platform, not as an option on an enterprise tier.

There is also a named customer environment. Moveworks GovCloud is described as a dedicated, secure environment built for government agencies and federal clients, backed by FedRAMP authorization. So both a named region list and a named customer environment are in place. On premises deployment is not documented.

Sourcemoveworks.com/us/en/platform/security and /solutions/ai-for-federal-governmentread 2026-09-12

An AI Agent Marketplace offers prebuilt agents that are free to install, and a studio lets customers build more. The marketplace has its own host at marketplace.moveworks.com, sits in the main navigation and is described as "a curated library of ready-to-use, pre-built, production-proof AI agents that are customizable and free to install, providing quick connector setup and MCP." It is a catalog to browse whose units do work once selected, with a stated install path and no added license, and each agent installs on its own, so removing one leaves the rest whole.

Next to the catalog is a builder: Agent Studio "enables developers to easily build and deploy AI agents that plan, reason, and securely execute actions across enterprise systems," and Knowledge Studio, Quick GPT and Brief Me ship as separately named units. Specialized AI Assistants, which create expert agents, are marked "coming soon" on the AI Assistant page. The marketplace may show how large the catalog is and whether entries come from Moveworks or partners.

Sourcemoveworks.com/us/en/platform/ai-assistant, /platform/ai-agent-builder and marketplace.moveworks.comread 2026-09-12

Employees are reached without asking, through nudges and broadcasts, on web, chat and mobile among other surfaces, in over 100 languages. The proactive side is a stated product behavior: the AI Assistant "anticipates what's next, sending proactive nudges for recurring or time-sensitive tasks." A second path ships as its own capability: Employee comms lets internal communications teams "broadcast timely, targeted comms to all or a subset of users so they never miss an important reminder to take action," a scheduled send to a chosen audience rather than a reply.

The channels are broad and listed. Employees can "access the AI Assistant on web, chat, or company portals," and the platform page adds intranets, browsers and mobile, which the vendor sums up as "omnichannel and multilingual" and "AI that works where employees do."

The Conversations API lets the Assistant be embedded in applications Moveworks does not own, support covers "100+ languages," and a dedicated AI Assistant on web ships as its own release. The Reasoning Engine acts on requests from any of these surfaces, and nudges key on recurring and time sensitive conditions rather than on a person starting a run.

Sourcemoveworks.com/us/en/platform/ai-assistant, /platform and /platform/internal-communicationread 2026-09-12

Several language models are routed on the customer's behalf, but no provider is named and no choice is offered. The Reasoning Engine is described as an agentic AI system that "orchestrates enterprise-ready LLMs and built-in tools for AI agents," so the routing is plural, done by the vendor and disclosed. The trust page backs up a multi model estate with safeguards applied "to protect LLMs" against hallucination, data poisoning and prompt injection, and commits that "no customer data is used to train global generative models." Moveworks historically ran on its own reasoning stack.

The customer chooses nothing. There is no model picker, admin setting or bring your own key path, no model assigned per agent, and no individual provider named. Orchestration is Moveworks' decision, made per task, and the vendor presents that as a feature, value on day one with "no model tuning, prompt chaining or advanced engineering required." Support for any model on the ServiceNow AI Platform belongs to the acquirer. The reasoning engine product page may describe a way to select models.

Sourcemoveworks.com/us/en/platform, /platform/ai-assistant and /platform/securityread 2026-09-12

Outside applications can call the Assistant through the Conversations API, which is backed by a developer estate, a build studio and shipped MCP. The AI Assistant page names the Conversations API as the way to "extend the Assistant even further," so Moveworks can be embedded in another application, portal or intranet.

The developer estate sits on four hosts in the main navigation and footer: a Developer Hub at moveworks.com/us/en/developers and help documentation at help.moveworks.com, which includes a Moveworks Setup admin reference, plus a community at community.moveworks.com and training at academy.moveworks.com.

Agent Studio is a build surface for the platform itself, where developers build and deploy agents and plugins that act across enterprise systems, and the Assistant is "linked to thousands of action plugins." MCP ships as well, named on the AI Agent Marketplace as part of quick connector setup, so an outside assistant can reach Moveworks. The Conversations API appears on a product page rather than in a reference, and the Developer Hub and help.moveworks.com may hold its endpoint list, authentication scheme and SDK.

Sourcemoveworks.com/us/en/platform/ai-assistant, /platform/ai-agent-builder and the developer estate linked in the site navigation; moveworks.com/us/en/developersread 2026-09-12

Output is checked for accuracy at run time and admins get tools to tune the Assistant, but no way to test a change before employees see it is documented. The trust page lists "fact verification to optimize accuracy of information generated" as a standing safeguard, alongside query risk assessment and a content moderation filter, which check what the agent produced.

Around it sits a tuning loop aimed at the Assistant itself: Employee Experience Insights gives program owners tools to "identify issues, track metrics, and enhance your assistant's efficacy," and the platform page lists "cutting-edge optimization tools" where "admins can easily track metrics, resolve issues, and make enhancements."

Nothing puts a change under test. No test environment, sandbox, holdout or simulation is documented, and neither is versioning or a scored result to compare. Fact verification improves accuracy at run time but gives the customer no verdict to compare across versions, and no page says a new or edited agent is run against a known set before it reaches employees, though these agents reset access and provision tools. The Employee Experience Insights page or the Agent Studio page may describe a harness used before deployment.

Sourcemoveworks.com/us/en/platform/security, /platform and /platform/ai-assistantread 2026-09-12

The browser is a place the Assistant lives, not an interface it drives, and no browser or computer use is documented. Browsers appear among the access surfaces, "AI that works where employees do: chat, browsers, intranets, service portals," and an AI Assistant on web ships as a destination employees open. On every documented path the employee navigates: the person opens the surface and types, and the Assistant answers and acts.

Actions run through code. The Assistant is "linked to thousands of action plugins that turn intent into results," and the Reasoning Engine "selects the right tools" to reach the goal across hundreds of connected systems.

Resetting access, provisioning tools, updating CRM records and submitting purchase requests all run through those connectors and the Conversations API; nothing clicks, fills in or advances a page, so a change in a target application's markup has nothing to break.

No hosted or local browser session, desktop control, remote or local computer use, or RPA fallback appears on any of the platform pages.

Sourcemoveworks.com/us/en/platform/ai-assistant, /platform and /platform/ai-agent-builderread 2026-09-12