Recently active topics

🎉 Update: MLS for Files is live as of May 18, MLS for all customers in all regions! You can now ingest non-English files such as PDFs, MS Word and MS PowerPoint and search will work the same way it does for KBs. See post below for full details. We're excited to announce more powerful, upgraded Multilingual Support (MLS) capabilities coming to Assistant and Enterprise Search!Today, via the Knowledge Search (KS) plugin, Moveworks supports non-English search for Knowledge Base Articles (KBAs) by converting both non-English queries and non-English KBAs to English in order to process them. While this approach provided multi-language support, it introduced risk of accuracy loss along the way. Additionally, non-English search is currently not supported for files.With our upgraded, native MLS capabilities, we will support non-English search for files along with KBAs. We’ll eliminate the "double translation" approach entirely, leading to improved accuracy and reliability. Our new upgraded sear

Hi everyone,I am looking for input on how other teams handle plugin testing and lifecycle management. I'd appreciate hearing what is working well for you!Our current process:We build and test in our Sandbox environment. We temporarily add testers, and once testing passes, we remove them to keep the environment clean and restrict access. We then export the plugin to Production and run a quick smoke test to confirm it works. For enhancements: We build the changes in Sandbox, export the new version to Production, and then swap them (turn off the old Production plugin and turn on the updated one). We have to keep the original plugin active in Production while we build the enhancement in Sandbox. What we are trying to figure out:We are looking for a simpler approach. We considered just building directly in Production and using the Launch Configuration to whitelist our testers (via "Allow selected users"), but we hit two roadblocks:System Triggers (Ambient Agents):How do you safely test webh

I tried submitting request which requires attachment to Workday, and according to the documentation, the Content-Type needs to be “multipart/form-data”: I tried via Postman and it works properly:Now, my problem is:How to create a HTTP Action in Agent Studio for this “multipart/form-data”?Specifically, how to translate the <Key, Value> pair in the “form-data” section in Postman, to the “Body” section of the HTTP Action Editor in Agent Studio?  I searched and found a related thread, tried passing the <Key, Value> as JSON but doesn’t work: May I know if HTTP Action Editor in Agent Studio is capable of such content type yet?Thank you.

Are there QA environments for Moveworks? If so how do I know I have enabled Moveworks QAs instances on my end? Currently I’m working with our Production Moveworks instance and it is a huge lame...

We're excited to share that we are beginning to support GPT-5.4 in the Moveworks Reasoning Engine. We put GPT-5.4 through extensive internal testing before bringing it to you.Stronger Instruction Following. GPT-5.4 adheres more closely to the instructions that shape your plugin's behavior, resulting in more consistent, predictable outputs aligned to your configuration.Improved Cited Answers. When your assistant responds with information from your knowledge base, it now includes links back to the source documents far more consistently. Users get answers they can verify, not just bare claims.Improved Tool-Call Argument Filling. When your assistant calls a tool, it more accurately extracts the right information from the conversation. Whether it's a ticket title, a PTO date, or an employee ID, GPT-5.4 more reliably finds the particular values it needs.Why we’re so methodical with the upgradesBefore rolling out any new model, our team runs it against a suite of internal benchmarks. We then

OpenAI recently released GPT 5.2, and we're excited to share that we've been hard at work evaluating it before bringing it to you. We ran a full assessment across a range of the latest large language models and put GPT 5.2 through extensive internal employee testing.More comprehensive tool calling. GPT 5.2 is more comprehensive in the sets of tools it uses to help answer user questions. As a result, it provides a more helpful answer, and is less likely to need to handoff users to internal help channels.Better Enterprise Search. GPT 5.2 works better with Enterprise Search. GPT 5.2 provides a more concise and relevant search query when using Search. This means the assistant finds what you're looking for more efficiently, with less noise in the search process.Stronger Tool Call Post-Processing. After a tool or plugin is called by GPT 5.2, it does a better job of writing code to analyze and restructure the results.Less Hallucinations. GPT-5.2 does a better job of discerning what capabiliti

When installing plugins from the Moveworks Marketplace, we've observed that certain plugin components do not get installed or appear in Agent Studio after the installation completes. This results in incomplete plugin functionality and broken workflows.Problem SummaryIssue Type: Incomplete Plugin InstallationSource: Moveworks MarketplaceAffected Area: Agent StudioImpact: Missing components, reduced plugin functionality

We are experiencing a recurring error when working with Conversational Processes in Moveworks. The error prevents the process from being published, which subsequently blocks us from using these processes in plugins.Error Details Field Value Error Message Sorry we cannot process your request this time Error Code 13121 Impact Conversational Process cannot be published Downstream Effect Unable to use affected processes in plugins

When installing plugins from the Moveworks Marketplace, we're encountering Error Code 13121 which prevents successful installation. The error indicates a missing value in Workflow Configurations > Workflow Arguments.Problem SummaryIssue Type: Plugin Installation FailureSource: Moveworks MarketplaceAffected Area: Agent Studio / Plugin WorkspaceError Code: 13121Impact: Plugin cannot be installed, workflows remain incompleteError DetailsError Message: There are errors with the field: Workflow Configurations > Workflow Arguments [arguments[0].type_id] - value must be setError Code: 13121Trace ID: qRkjwu_cbqsTWU-U 

This is related to our ITSM process in ServiceNow. I’m looking to trigger a notification by our bot on Teams when the State of a ticket is changed, like an event-driven action. I noticed that there is an existing notification trigger but this only works when the fulfiller has written down Additional Comments (Customer Visible), otherwise it does not trigger any notification. 

Was really excited about ‘Vetted Content’ being released (https://help.moveworks.com/ai-assistant/enterprise-search/vetted-content) however even after adding a few articles across different systems (confluence - next gen connector and google drive - regular connector) docs are showing as Available but not highlighted as vetted sources.The knowledge returned is correct, and its returning the correct knowledge articles, but based on “What end users see” https://help.moveworks.com/ai-assistant/enterprise-search/vetted-content#what-end-users-see it should be flagged. Seeing similar behavior between the slack and web assistant interface. Opened a support ticket, but figured it would be faster to poll the community than get an answer from support. 

 6/4 update: Vetted Content is now live on Assistant (Web, Slack and Teams) in addition to the Enterprise Search Web App! Thank you for your patience as we worked through the remaining items — we're excited for you to try it out. We’re launching a new feature called Vetted Content which will be generally available (GA) on May 28, 2026 on Enterprise Search Web App to help employees instantly recognize approved, authoritative answers in Moveworks Assistant and Enterprise Search. Please note: this update is designed to work best with GPT-5.2, currently in our Frontier and Standard plans (link). Under GPT-5.1, Vetted Answers may not perform optimally.What it isVetted Content clearly signals which indexed files and Knowledge Base Articles (KBAs) are approved by your organization, so employees can quickly trust the answers they see. Use Vetted Content to improve trust when your organization contains outdated or unmaintained content, multiple documents cover similar topics and employees can't

Hey,I am reaching out to check if anyone has had success with the GitHub ingestion setup outlined in this documentation: https://docs.moveworks.com/ai-assistant/enterprise-search/configure-search/ingestion/how-to-configure-github-knowledge-ingestion.I configured the connector and ingestion as per the article. The records are successfully being ingested from the repo, but the base URL is being dropped and the .md suffix is being stripped, resulting in broken links (System doesn’t allow me to leave “Github Config Article Url Filepath Suffixes To Drop” as blank).For example, src/ad/README.md redirects to https://src/ad/README. The full path should actually be: https://github.com/newrelic/opentelemetry-demo/blob/main/src/ad/README.mdFor context, my ingestion config Start URL is: https://api.github.com/repos/newrelic/opentelemetry-demo/contents/Am I missing a step in the configuration to retain the base URL and the file extension? 

We are experiencing an issue with HTTP Connectors in Moveworks where, once a connector is created and saved, we are unable to modify or update its configuration. After successfully creating an HTTP Connector, the connector becomes effectively read-only from an administration perspective. We are unable to edit existing configuration details such as:Base URL / Endpoint URL Authentication settings Headers Query Parameters Request Body configuration Connector metadata and settingsAs a result, whenever a configuration change is required, we are forced to create an entirely new connector instead of updating the existing one. Business ImpactThis behavior creates several operational challenges:Increased administrative overhead Duplicate connectors in the environment Difficulty maintaining connector versions Additional effort during testing and troubleshooting Potential risk of workflows referencing outdated connectors

I am attempting to do the following:Loop through a list of teams and for each one run an HTTP action to get the list of members and for each of those members run another HTTP action to get their emails.I want to map the output so that i get the following:[{ “team_name”: “Team A”, “emails”: [“email1@email.com”, “email2@email.com”] },{ “team_name”: “Team B”, “emails”: [“email3@email.com”] }]I tried going through the agent architect and its saying to use data_mapper: but that doesn’t seem to be allowed in a compound action Is the best way to achieve this a script action using python to manipulate the outputs?Not sure why the architect is suggesting something that is unavailable for use. 

Hi MW Community,New to MoveWorks and I want to understand if a use case I am exploring is possible. A simple use case example:Open the assistant with the context that my favorite color is blue. This will need to happen behind the scenes without me explicitly telling it. Assistant able to answer ‘what is my favorite color’ by referencing the context.Some open questions:  1. Does the Moveworks embed/web SDK accept arbitrary context on init that becomes available to the reasoning engine?                                                                                              2. Is there a "conversation starter" or "welcome plugin" concept that fires automatically on open?                               3. Is there a way to initiate a plugin to read from URL query parameters on open conversation? Thank You!

I cannot download the User Feedback column from Moveworks Analytics > AI Insights > Interactions Table > Raw Interactions. The column displays on screen but does not download as a CSV even when all three options are selected.

I am working on a ambient agent that is able to send notifications to 1 or more users. The listener is expecting a payload with an object that has email and the message to send to the user. I wanted to know if there is a size limitation on the API call coming into the listener. Trying to determine how much to limit the bulk api call size if needed. 

Hi everyone,Our organization is preparing to connect to the Moveworks Data API to pull metrics into our own BI tools to better measure the overall value of our bot. Before we kick off the technical work and internal security reviews, I would love to hear from other customers who have already gone through this process.We are currently trying to map out our internal dependencies and understand the real-world effort. If you have successfully set this up, I would appreciate your insights on the following:1. Setup & Implementation EffortWhich internal teams did you need to involve (e.g., Data Engineering, InfoSec, Network)? What did the setup work look like on your end versus what Moveworks handled? How long did it take from the initial request to your first successful data pull, and what were the biggest blockers?2. InfoSec & Network ValidationDid your InfoSec team have any major concerns regarding where the data is stored or the endpoints/URLs required? Is the API exposing anythin