New Moveworks User: Getting My Bearings with Automation Tools

hey ​@pierce_neptune! sorry for the late reply, i’d like to preface that i’m not an expert here and also am still working through what i’m about to suggest 😅.

1. in our experience, the thing that provides the most immediate value is the ticket concierge/triage combo, if you can get those rolled out (or already have them rolled out), that’s the biggest ROI in my opinion straight out the box.
2. the next thing i would focus on for sheer value is building governance rules and processes around your Knowledge Base (SNOW, Confluence, self-hosted website, etc.) and fine tune your ingestion practices. the reason for this is because most people’s first interaction with the bot is going to involve the bot serving or looking for pre-existing knowledge to answer their question/solve their request. people’s perception of the bot (again, in my exp) is almost entirely driven by their first impression of whether the knowledge it served was helpful. so if your knowledge base solution is lacking, user’s will think the bot is lacking, which isn’t necessarily true.
3. after building a strong foundation there, i would definitely start diving into Agent Studio automations. there are definitely some “quick wins” opportunities like: calendar scheduling, password or auth resets (work pro-actively with your sec team on anything IdP related), pto requests, etc. the key for AS for me has been learning and grasping: Compound Actions syntax/Moveworks DSL, experimenting with Slots and Resolver strategies, and filtering your API calls to include only the data you want to receive in your payloads.
4. i’m not sure if you have access or not, but if i was starting over with AS, i would deep dive and read through the DSL docs for a week while attempting to build actions/Conversational Processes. once i felt like i had the basics, i would start leveraging Agent Architect to check my work and ask more advanced format and solution questions

These are mainly regular reports that I send on a frequent basis, requiring downloading information from a few sources and performing some basic analyses.

I can’t tell if my tasks are ideal “use cases” for the scope of Moveworks automations, or if I should handle that kind of individual automation in a different way that makes more sense for an individual contributor task, focusing instead of Moveworks automation for tools that are intended for a broader group of colleagues?

To speak directly to this - it sounds like this work is certainly right up an AS use case. if there is an API open to download your data from, you could setup an HTTP call to the endpoint, then with the data in the return payload, write a script action to process the data how you want, then finally return the data back to you in whatever format you’re looking for. this would definitely save you time, however, the key i believe is simply doing the math on how many minutes/hours this will save you in a given month/quarter for example by having the bot take care of it. check that against the cost of publishing the use case and there’s your answer, at least that’s how we tackle which use case we are going to prioritize.

once you get the hang of developing the use cases, you’ll quickly begin to identify tons of opportunities in your org to save users time. as long as the number of hours returned to the user outweighs the cost of the use case and man hours it will take to develop, you’re golden :)

hope some/any of this helps! feel free to respond with any more questions, in the weeds or not i’m happy to chat.