Build better products with our product team.
We are working on significantly improving the relevant search results that are surfaced to users
After surfacing search results, the bot will sometimes suggest follow-up actions that it’s not capable of. We are working to improve the bot so that it only recommends actions that it is capable of.
Sometimes the bot says it completed an action when it really didn’t. We are working on giving the bot the context it needs to always surface truthful information to the user.
We’ve observed scenarios where the bot will ignore user’s attempts to update slot values after that specified has already been filled. We’ve fixed an issue that previously prohibited the bot from doing this.
Slot filling behavior has been inconsistent — in some cases, slots aggressively infer values that were never explicitly provided by the user, while in others, they fail to populate even when the user clearly stated the value.We are aiming to improve the overall reliability of slot inferring across the board.
We had captured multiple scenarios where the original article of a search result would link out to some resource (portal, form, additional info) but the bot would not surface these helpful links to the end user.Now, we persist hyperlinks that meet the following criteria:Surface hyperlinks when:A specific artifact (form, guide, doc) is referenced and a deep link exists in retrieved content The user needs to navigate somewhere or complete a workflow that can't be done inlineDon't surface hyperlinks when:The link is only a generic homepage or broad portal The artifact is mentioned conceptually with no navigation needed The workflow can be completed fully inlineIf multiple links exist: always prefer specific deep links over generic/root-domain URLs.
In Agent Studio, Activities have a “Confirmation Policy”. This requires the user to explicitly give the bot go ahead to execute the action. However, we are seeing that the bot sometimes skips this step, even when the Confirmation Policy=True.
Developers will be able to configure citations for Action Activities, letting users verify the data behind every action.
Upload a custom AI Assistant avatar directly in Moveworks Setup that will appear across all user-facing experiences, including Moveworks Web, popup forms, ticket links and more.
Upload bot avatars directly in Moveworks Setup, eliminating the need to host images and provide image links
Analyze ticket data in the Moveworks ticket object format to uncover issues in ticket mapping and concierge skill execution.
Troubleshoot content serving scenarios by evaluating if an employee will be served any specific ingested content or not.
Servicenow Approvals are a high value skill for many customers that would be amplified in value if expanded to support the delegation functionality.
I think the concept around the Employee communications module is great but many small tweaks could make it so much simpler and better to use. Importing users -Why not allow me to select from users? Uploading a CSV is fine but if I want to send to like 20, why not just allow me to select them from the data you have.Running Campaigns - Analytics leaves a lot to be desired - Downloading a spreadsheet is painful Why can’t I have a table view of the charts? Why not allow me to click the bar charts and get deeper into the data? I have a user who clicked the button 5 times, let me send a different message to them asking if they need help or something. Buttons/Actions - It feels disconnected from the LLM If one option is to open an incident or a form that the bot can serve, why do I only have the option to link out to my ITSM? Why not have the experience stay within the bot? Add a button call serve Incident form, Serve X, Y, Z request form. Allow for treeing of responses like a Survey If you select option A, you get two further choices.
Need MW to allow submitting Thumbs up/positive feedback directly without the need to open a pop up window and then submit. This will simplify the process for capturing positive user sentiment. Currently users have the follow the standard flow for additional clicks. We need to remove additional clicks and upon click of thumbs up bot could accepts it as a positive feedback. thumbs down can follow the below standard path as its crucial to get the reason for negative feedback to help improve. click on thumbs up - opening popup, add comment in the text field (optional) click on submitRequire users to provide mandatory text feedback when giving a thumbs down feedback. This will help us understand the reasons behind the negative feedback and improve the bot accordingly.
In the Moveworks Gov environment, we have a need to be able to ingest the Configuration and Permissions logs into our corporate SIEM. Currently, those logs are only available in the Moveworks Console. As part of our CMMC requirements to authorize Moveworks to handle CUI data, we need to have a centralized location for all of our logs to be ingested and monitored by our SOC. We are already ingesting the logs that are available to us for API calls but adding a mechanism to also be able to ingest these logs into an external system is likely a requirement for many other CMMC accredited companies. As an example, log entries such as someone being elevated to a Super Admin would be the type of activity that our SOC needs to be able to monitor across all of our SaaS applications.
Already have an account? Login
Enter your E-mail address. We'll send you an e-mail with instructions to reset your password.