Genesys Cloud - Main

 View Only

Sign Up

Expand all | Collapse all

AI Guides - what do you like/dislike? Have you actually deployed any?

  • 1.  AI Guides - what do you like/dislike? Have you actually deployed any?

    Posted 09-01-2025 17:24

    Just curious, for those that have deployed or have at least toyed with the idea of deploying AI Guides, what challenges have you met and which ones prevented you from actually deploying it?  Aside from the financial/token expense that is :)


    #ConversationalAI(Bots,AgentAssist,etc.)

    ------------------------------
    Vaun McCarthy
    ------------------------------


  • 2.  RE: AI Guides - what do you like/dislike? Have you actually deployed any?
    Best Answer

    Posted 09-01-2025 23:29
    Edited by Jason Kleitz 09-02-2025 09:24
    Good questions, @Vaun McCarthy. First & foremost, we really appreciate the convenience AI Guides offers with respect to reducing prompt & flow scripting. That alone will provide huge time savings & increase our agility.
    Before sharing our issues/requests, I've been told 1) [for now] AI Guides best works in a hybrid fashion (i.e. Virtual Assistant supplemented with very simple AI Guides), and 2) AI Guides will be frequently updated to address limitations/issues/advancements (e.g. next update coming in a month or so). As such, our AI Guide (i.e. mostly Agentic experience) expectations have been lowered [for now] to a basic level to ensure what we do deploy (with some AI Guides sprinkled in) is as successful as it can be under the current circumstances. With that in mind, here are our top issues & enhancement requests:
    -----------------------------------
    TOP ISSUES (Note: We have open support Cases for these and have been told they should be addressed in the next Guides release):
    • Issue 1: Not Consistently Following Prompt/Guide Instructions
    • Description: Guides are not always following specific instructions, specifically those that tell them when not to do something. Not to mention the Guides can tell you about their prompt and what step they're on unless you explicitly tell them not to. Furthermore, if you correct the guide based on the instructions, it will admit when it make a mistake in the instructions and correct itself. Ironic how "deterministic/explicit" we still have to be for what's supposed to be a non-deterministic experience. Also interesting how these types of nuances are also not included in the resource center documentation, thus resulting in onoing, unexpected trial and error.
    • Impact: Unable to rely on Guides consistently following the desired instructions. We have tried modifying the instructions multiple times to account for these gaps and even had Genesys SME input that hasn't yielded the desired results.
    -----------------------------------
    • Issue 2: Stalling/Unnecessarily Waiting for User Input
    • Description: Guides occasionally require the user to say something before it proceeds to the next step; i.e. the prompt clearly states to do X and then Y, but it sometimes does X, but doesn't do Y until the user responds. It seems unnecessary to constantly ask the user an arbitrary question after each step or spoken phrase so the user is forced to respond in order for the Guide to proceed to do Y. This is not how normal communication works, especially if the Guide is used in a voice channel. For example, the times when the Guide appears to listen to the step to return to Architect, it hangs at the last communication until the user responds. Once the user responds, the Guide responds and then transfers the interaction back to Architect.
    • Impact: Unable to rely on the Guide to hold a normal conversation without causing customer frustration/uncertainty.
    -----------------------------------
    • Issue 3: Data Actions Not Consistently Working
    • Description: We have written our prompts per the Resource Center documentation recommendations and have modified them multiple times per Genesys SME recommendations, yet the Guide fails to consistently call the desired data actions.
    • Impact: Unable to rely on Guides consistently performing the desired tasks.

    -----------------------------------

    ENHANCEMENT REQUESTS:
    • Idea SAAIS-I-22 | 5 data action limit is too low/restrictive. For now may continue to perform major data actions outside the Guide and feed the results into the Guide as variables. Problem there is we only get 5 variables.
    • Idea SAAIS-I-24 | Data action output variables that are used in the prompt need to be underlined or highlighted so they stand out.
    • Idea SSAIS-I-25 | Given you cannot rename, add, or delete variables once you publish with 5 used variables, it would be ideal if you could add as many as you want and toggle which 5 are active.
    • Idea SSAIS-I-26 | Cannot rename Guides.
    • Idea SSAIS-I-27 | Need data table lookup feature.
    • Idea SSAIS-I-28 | Need Knowledgebase article lookup feature.
    • Idea SSAIS-I-29 | Execution Flow History needs to show Guide steps & data.
    • Idea SSAIS-I-30 | Need version history like with Architect flows

    -----------------------------------

    P.S. Did you know you can ask the Guide to communicate like a pirate, and it will? Why customers might do that, I don't know, but figured you might want to know just in case it could pose a problem/risk. I'm afraid to ask it to take on any other personas though, haha-hargh 🏴‍☠️



    ------------------------------
    Brian T. Jones | Ascension | Senior Specialist - Technology
    ------------------------------



  • 3.  RE: AI Guides - what do you like/dislike? Have you actually deployed any?

    Posted 09-02-2025 00:45

    Brilliant, thanks Brian. That's exactly the type of response I was after.  Interestingly but perhaps not surprisingly the 3 issues you list there essentially mirrors my own challenges.

    I know that the resource centre does say:

    • Ensure that every step in the guide makes your AI agent anticipate a customer action or response.
    • If the current step has no such query, combine it with the next step that does require a customer reply. 

    But as you point out that's fairly restrictive - especially in the case of something like "is there anything else I can help you with" which is handled back in the main bot flow in some cases.  You have it do something like give an account balance, and want it to immediately just go back to the bot flow but instead it seems to get stuck waiting on a response to a question that was never asked :)

    I did raise with our CSM and TAM the fact that on occasion you could convince the guide to tell you the variables in use, including the current value, and tell it to update a particular variable with a different value.  It doesn't seem efficient to have to throw all that into "guard rails".

    Which leads me to another part - trying to understand the correct syntax/order/heading to put guard rails in the guide.  Some Genesys videos and demonstrations I've seen have this in different formats but it would be useful to have a clear definition or "guidance" of how that should be included in the guide's instructions.

    It's an absolutely great idea which could be brilliant if there's some expediency with enhancements - and not just sitting waiting on votes or community feedback on the ideas portal :)

    On paper it seems very nice but that niceness loses some points with the amount of tuning you have to do in a guide's instructions after it's generated.



    ------------------------------
    Vaun McCarthy
    ------------------------------



  • 4.  RE: AI Guides - what do you like/dislike? Have you actually deployed any?

    Posted 09-02-2025 05:52
    Edited by Brian Jones 09-02-2025 10:20

    Well said, Vaun. At the moment there's certainly a lot to say on this topic, however, until Guides evolves there will be more constructive criticism than outright praise.  As Mitch said when announcing this, their approach is "fail fast & fix fast", which is coincidentally what our team has had to do, haha

    Right now the primary thing we love about them (the non-detrministic side) proves too inconsistent & risky almost to the point of driving me crazy. Ironically, after field testing we're having to include just as many instructions in the Guide regarding what not to do as there are instructions on what to do (which, to your point, isn't included in the resource documentation & such.) By the time we're done massaging a Guides prompt, we could've built an equally effective deterministic virtual agent experience 🤷

    All that said & to your point about the account balance, we've been informed to resort to the hybrid model I mentioned,  which makes sense in theory, but is still plagued with inconsistency given (by nature) the Guide truly wants to be "autonomously conversational"--thus pushing the limits of what it can do/say based on what it was not told to not do/say. As such, we end up resorting to using the Guides to do simple data collection in a conversational manner (yay--no TTS/audio prompts, no intent/slot input/matching configs, or error handling) and the virtual agent to do the processing/communicating of said information--all at the expense of 1) having to program numerous Virtual Agent > Guide blocks (to do "Guide switching") and 2) hoping the Guide consistently stays in its lane.

    I'm sure AI Guides will get there--wherever "there" is, relatively speaking. In the meantime, it's certainly a lot to think about & stay on top of.



    ------------------------------
    Brian T. Jones | Ascension | Senior Specialist - Technology
    ------------------------------



  • 5.  RE: AI Guides - what do you like/dislike? Have you actually deployed any?

    Posted 09-02-2025 11:21

    Hi all,

    Our team has ran into much of the same issues described above and would love to keep track of progress toward addressing the current limitations / issues. The AI Guide offering has really great potential and with a few tweaks, seems like the offering can go from good to great. One other Idea that we submitted that was denied is to make AI Guides division aware (or at least place some security constraints around who can view/edit Guides). 

    https://genesyscloud.ideas.aha.io/ideas/SSAAIS-I-31 - AI Guides: Division Aware

    With the goal/intent of these guides to push down to each line of business, it really seems logical to be able to restrict who can view/edit certain guides based upon Division or specific conditions. We have multiple divisions and there is a chance we'd have competing lines of business each in their own division. They most definitely would prefer to have their guides only viewed by their own staff. 

    Ideally for us, the solution would be AI Guides that are division aware with both a unique Edit and View permission. Some of the staff would only need to View the current iteration of the guide.

    Will definitely follow this thread and hopefully learn more about next steps within the thread or in Nashville next week.

    Great thread! Thanks all! 



    ------------------------------
    Shane Jenkins
    ------------------------------



  • 6.  RE: AI Guides - what do you like/dislike? Have you actually deployed any?

    Posted 09-02-2025 11:28
    Edited by Brandon Weaver 09-02-2025 16:26

    I've spent the last week trying to create a demo video for the the Virtual Agent, Guides and Copilot. The idea was to showcase these features to current and potential customers.

    However, Guides in particular have given us pause.

    I've experienced everything @Brian Jones listed above numerous times. So far, no amount of guardrails or "massaging" the Instructions has helped to eliminate these issues.

    I'll simply echo Brian's words here as they capture our frustrations quite nicely:

    "Right now the primary thing we love about them (the non-deterministic side) proves too inconsistent & risky almost to the point of driving me crazy. Ironically, after field testing we're having to include just as many instructions in the Guide regarding what not to do as there are instructions on what to do (which, to your point, isn't included in the resource documentation & such.)
    By the time we're done massaging a Guides prompt, we could've built an equally effective deterministic virtual agent experience."

    For now, if we decide to deploy Guides at all, we'll be using them for the simplest of tasks. 
    For anything more, we'll be doing a lot more in Architect - outside of the Guide - than we had originally hoped or thought necessary. 

    I still think Guides are really cool and I hope to see them improve. I can imagine us using them quite heavily once they become more reliable.



    ------------------------------
    R. Brandon Weaver
    Science Applications International Corporation (SAIC)
    ------------------------------



  • 7.  RE: AI Guides - what do you like/dislike? Have you actually deployed any?

    Posted 09-02-2025 19:10

    Jones, I noticed an oddity with my own Guide where after I got it to talk like a pirate a few times, it started refusing to do so. I haven't gotten around to asking internally yet whether they changed something, or if it somehow "learned" it shouldn't do that...

    Can you still reproduce at will?



    ------------------------------
    George Ganahl GCCX-AI, GCP, GCSME
    Technical Adoption Champion
    Genesys
    2024 Community Member of the Year
    ------------------------------



  • 8.  RE: AI Guides - what do you like/dislike? Have you actually deployed any?

    Posted 09-02-2025 19:11

    Also...looking forward to seeing you at Xperience!!!



    ------------------------------
    George Ganahl GCCX-AI, GCP, GCSME
    Technical Adoption Champion
    Genesys
    2024 Community Member of the Year
    ------------------------------



  • 9.  RE: AI Guides - what do you like/dislike? Have you actually deployed any?

    Posted 09-03-2025 13:59

    I just noticed that as well, George. Looks like they updated the model on the backend to not take on any other personas, haha. I'll let you know if it eventually allows me to impersonate anyone else.

    See you at Xperience!



    ------------------------------
    Brian T. Jones | Ascension | Senior Specialist - Technology
    ------------------------------



  • 10.  RE: AI Guides - what do you like/dislike? Have you actually deployed any?

    Posted 09-03-2025 18:19

    yessir...Monday morning 



    ------------------------------
    George Ganahl GCCX-AI, GCP, GCSME
    Technical Adoption Champion
    Genesys
    2024 Community Member of the Year
    ------------------------------



  • 11.  RE: AI Guides - what do you like/dislike? Have you actually deployed any?

    Posted 09-02-2025 19:09

    I am definitely saving this thread...



    ------------------------------
    George Ganahl GCCX-AI, GCP, GCSME
    Technical Adoption Champion
    Genesys
    2024 Community Member of the Year
    ------------------------------



  • 12.  RE: AI Guides - what do you like/dislike? Have you actually deployed any?

    Posted 09-09-2025 09:15
    Edited by Lionel Florence 09-09-2025 09:18

    Thanks for opening this discussion - I agree with many of the points already raised and wanted to share some feedback based on our own deployment experience with AI Guides.

    Documentation & Implementation
    The implementation documentation is incomplete and far too limited. It lacks depth and doesn't provide enough guidance for real-world scenarios or troubleshooting.

    Inconsistent Behavior
    The behavior of the guides is often unpredictable. For example, when configuring a question to be asked three times, it usually works - but not always. Sometimes the guide only asks twice, without any clear reason.

    Flow Construction
    The Inbound Message Flow → Digital Bot Flow → Guide structure is solid. However, when the guide fails to understand the input, the exit behavior is inconsistent: sometimes it triggers SUCCESS, other times Recognition Failure, even though Failure or Timeout would be more appropriate.

    Data Validation Issues
    We frequently encounter issues with input validation. For instance, validating an ID in the format S###### (where # is a digit) works most of the time, but fails randomly.

    Fast User Input Confusion
    When users respond too quickly, the guide tends to mix up questions and answers, leading to confusion and often resulting in escalation to a live agent.

    Poor Readability of Guide Output
    I haven't found a way to improve the readability of guide responses. When multiple lines of information or questions are presented, they appear as a single block of text, which becomes hard to read and understand.



    ------------------------------
    Lionel Florence
    Directeur UDJ
    ------------------------------



  • 13.  RE: AI Guides - what do you like/dislike? Have you actually deployed any?

    Posted 10-22-2025 18:00

    Thanks Lionel, just on that readability part - I had an instruction in my guide to format this type of thing nicely.   But like other things in the Guides, I found it did that sometimes for one or two interactions, but just kept putting it out like you see in your example.  I've tried different wordings in the instructions to try to tell it to always format it like that but I still couldn't get consistent behaviour.



    ------------------------------
    Vaun McCarthy
    ------------------------------



  • 14.  RE: AI Guides - what do you like/dislike? Have you actually deployed any?

    Posted 09-16-2025 00:34
    Edited by Glen Tylee 09-16-2025 00:35

    I'm going through this now.

    Just trying to understand the limitations and have found some things that stand out as opportunities for it to fail, and not really in the way it responds so much, but in how it works overall. 

    • Character Limit: We have a very complex business so having a guide is great, but I easily reach a character limit while only being half way through the information that needs to be available to answer queries. We have strong regulatory requirements for a food product so can't skip on details. Under the current Digital Bot Flow I just have digital menus that lead to answers, which I'd like to move in the direction of being able to use AI to write responses, but if I can't input all the information that is required to accurately answer a question, then I have to revert back to using it as a navigation tool instead, which is what our Digital bot flow already does. So am I going to deploy AI just for increased navigation and not actually resolving a query...
    • Unable to move between guides: If a customer goes down the path leading to a guide on one topic, and then they decide to switch topics, I then have to ask them if they are switching topics or not. Maybe the information is just not in the guide but the topic is correct, or maybe they want to talk about something else. If they are wanting to change topics, then I have to have the Digital Bot Flow ask them what they want to talk about, which they have already asked while they were in the previous guide but couldn't get an answer to because of the character limit or the information is just not contained in that guide. If it fails to find the answer from that guide, it would be good if it could go and check other intents to see if there is a match before returning a 'no match' and then confirm the topic change.
    • No match: If the customer was communicating with the AI Guide, however as the conversation progressed the guide didn't contain the answer to the question, it would be nice if I could send them down a no-match path in architect to ask them if they were still talking about that topic. If they were (then I assume the guide does not contain the answer but the topic is correct) then escalate to an agent. I could then add information on the unanswered question later. 
    • Welcome Message on Guides: I ask the guide not to welcome the customer, because I already have a message welcoming the customer prior to entering the guide. So my welcome message say's 'Hello, in a few words what are you looking for today', then the customer types their answer, and then the guide responds 'Hello! I am here to help you with...'. If I tell the guide to not do a welcome, it more often than not, continues to do it. 
    • First message from customer just needs to be repeated: When people message us, they start with their question. For example they will start by saying 'where can I go to download XXXX', the AI responds by saying 'looks like your talking about Topic A', what can I help you with today'. So the customer has to re-enter the question they have already asked. Not ideal, especially if the question can't be answered by the guide either. Then it's two failures in a row before escalation. 

    I see how this might work for a company that needs to complete one or two tasks, like question an order status of a product or order a basic item. But anything more complicated than that, it's not there yet until more features come out from what I see.  
    I really want to use this but I think I need to see where this is going, as I think other products are already further along in this space. 
    I assume I am correct in all of the above, maybe there's solutions I am not aware of to this, but I can't see it at the moment. 



    ------------------------------
    Glen Tylee
    Systems Support Lead
    ------------------------------



  • 15.  RE: AI Guides - what do you like/dislike? Have you actually deployed any?

    Posted 09-22-2025 10:12
    Edited by James Dunn 09-22-2025 11:00

    I only started playing around with the guides recently but:

    1) I hate that there is a limit on variables. It makes it impractical when many of our Data Actions will spit out more than 5 things that we need to store and pass to the flow for the future. Workarounds for this are not good, e.g. if we use a Data Action in a guide to identify the client (lookup against CRM), we have them confirm it, then have pass the lookup variables out and repeat the lookup in the flow itself...

    2) I don't like that you can't edit / remove variables after publishing. 

    3) The guide can't save anything to Participant Data.

    4) Debugging is difficult. Going into the guide is like going into a black hole, you can't replay it, you can't step through it stage by stage etc. I was getting a strange response (and some hallucinations I suppose) from after a Data Action but I couldn't see what was actually causing the guide to act how it did.


    ------------------------------
    James Dunn
    Telecoms Specialist
    ------------------------------



  • 16.  RE: AI Guides - what do you like/dislike? Have you actually deployed any?

    Posted 10-15-2025 11:39

    hello,

    1) The layout of th eguide is not respected. For example, when in the guide we go to the line, in the bot the words are one after the other. The rendering is not optimal, it is not well laid out.

    2) I would like to have the option, in the guide, to choose to transfer the call directly to an agent. For example, say "transfer the call to queue xxx"

    Thank you :-) 



    ------------------------------
    Catherine DUPIRE
    Customer Success Manager
    NXO FRANCE SASU
    ------------------------------



  • 17.  RE: AI Guides - what do you like/dislike? Have you actually deployed any?

    Posted 10-22-2025 17:11
    Edited by Brian Jones 10-22-2025 17:15

    DISLIKE:

    So, I was excited to see the Preview Guides feature was added so we could test our Guides before publishing them (particularly so we could have generically named, multipurpose variables given you can't change them once published). However, I was a bit disappointed to see this requirement in order to run a data action within the Guide. Why?!

    Isn't the whole point of previewing the Guide for it to run an actual test, including the data actions? For example, our data actions (the same ones we're being asked to "manually provide values for") can output anywhere from 1 - 8 variables for each data action--and in some cases we have multiple data actions. This means the simulation stops at every data action asking us to add the output variable(s) before it can continue to the next step, which doesn't help us truly simulate an actual experience.

    Ideally there should be an option to run the data action using the provided inputs and the Guide to use whatever the actual data action output that was returned to continue to simulation. Off to the Ideas portal I go and back to using digital bots to test the Guides again🙄


    ------------------------------
    Brian T. Jones | Ascension | Senior Specialist - Technology
    ------------------------------



  • 18.  RE: AI Guides - what do you like/dislike? Have you actually deployed any?

    Posted 10-22-2025 17:54

    I agree Brian.  I was keen to see this preview function as well but that excitement pretty quickly fizzled out when I saw I wasn't able to actually preview it to test what I was trying to fix or achieve without still putting through a full interaction.  Given a lot of guides are being built specifically to do tasks that likely include data actions, not being able to simulate/test/preview directly within AI Studio still leaves a gap.  At best you can only preview/test some of your guardrails but a lot of things you've got no idea how it might operate without putting conversations through.  In which case what does the preview function actually give us?



    ------------------------------
    Vaun McCarthy
    ------------------------------