Genesys Cloud - Main

No worries @Rebecca Harper. 

Yes, Guru shared some guidance with me on how to achieve that. Thanks to Guru. I am looking for more in-depth information, such as whether it's possible to route to multiple queues if my Guide has several support scenarios that require routing to different support agents? Also, can a Guide link or call to another Guide?

Hi Hann Chiad Hooy, 

Forgive the late reply. I believe Guru replied to your email ont he same subject. I will discuss with him the idea of having another video covering this. For any other subscribers the advice given was to create an "intent" in the flow (example "escalate to Agent") that directs to a reusable task, and when the intent is hit in the Guide it will be redirected out of the Guide. 

This is one example of how it can be achieved. 

Hi Rebecca,

Thank you for sharing. Would you mind sharing the detailed steps on how to use the function to route the call to an agent/queue? (from Guide > Bot Flow) I just can't figure out what syntax/commands to use. If you have another video guide regarding this topic, that would be extremely helpful to me.

Hi Chris Dilworth,

I see your predicament. I expected that would be the problem, extending the timeout for the entire conversation isn't ideal. I haven't had a chance to test yet and I suspect it mightn't work, but have you tried adding "do not prompt the customer for a responce until 2 minutes of inactivity" in the Guide prompt? I feel like the inherited timeout wold take precedence, but it's worth a try. 

In the meantime, raise an Idea if there isn't one, as a lot of work is being done to improve later versions of Guides and being developed rapidly. You might get the functionality sooner than expected. 

Becky 

PS: If I manage to find time to test it, I'll update the thread with my outcomes or if I hear anything from internal teams on any workarounds. 

Thanks for the response, Becky.  Unfortunately, the max timeout in the bot flow is 10 seconds.  This would not help for two reasons.  First, we need an infinite timeout in which the customer instructs the guide to continue.  Second, if we set it to max in the bot flow, then the timeout for the bot's voice prompts would be incorrect.  We really need for the guide to have some controls that govern the behavior of the VA itself.  Maybe in another version...  Thanks.

Hi Chris Dilworth,

Thanks for reaching out. This can be configured in the BOT flow tat has the guide. You can set the no imput timeout for the whole flow. I don't see a way to make this dynamic based on the intent though, sorry. 

The gude(s) should inherit the setting from the BOT.

Hope this helps,

Becky 

We have noticed that the guide does not allow for a customer to complete a task that takes more than a few seconds before responding.  IE:  The guide has asked the customer to reboot a wi-fi router and come back when it is back in service.  The bot asks 3 times every 5 seconds if the task is completed and then immediately bails to a representative.  Where do we change this threshold, as many technical tasks take minutes at a time.   Thanks.

Hi Dan Weber, 

Thanks for your input. From my personal perspective I suspect you haven't done anything wrong in regards to configuration. For me, there is a noticable delay in voice responses and, if you have worked with Guides with digital media types, this reflects the period where the AI is computing and you typically see those three dots ... while a response is generating. 

Our product team is working hard to shave down the computing time as much as possible and, concurrently, identifying other ways to deal with this experience more gracefully. 

In the meantime, I hope that Product Support can help identify any areas that can be adjusted for improvmement. 

Kind regards

Becky 

We’ve observed that the AI Guide enforces a one second end of voice input timeout across all response fields, regardless of the configurations applied within bot flow actions. This behavior consistently leads to input failures when callers pause briefly while speaking. Whether they are providing a name, address, zip code, account number, or other details, even a short pause causes the system to prematurely interpret the input as complete, resulting in incomplete data capture and a disrupted experience.

We’ve attempted multiple adjustments, including changes to bot flow action settings and guide-level instructions, but none have resolved the issue. This specific area requires further attention and investigation. We have a ticket open with support.

Hi Heikki Tidenberg,

Thank you so much for your feedback. It is incredibly helpful to hear your experiences with the product. I have not tried your use-case yet and I am keen to give it a go. 

I will pass on your feedback to our product team so that they can work on improvements. 

Any time you have feedback like this, it is always welcomed. 

Kind regards,

Becky 

Hi!

I've been testing AI Studio Guide lately and I have used it to create an AI bot that acted as a customer service representative for a company. The bot was prompted to use the company's web pages as source material.

The bot worked great, but there were challenges in situations where the bot was asked to provide the URL of a product page. For some reason, the bot couldn't give the entire URL, but cut off the end of it.

Example:
Correct URL was: "https://www.company.fi/product-500-ml-101733"
The bot responded with the address: "https://www.company.fi/product-500-ml"
The bot cut off the "-101733" from the address.
I've tried to teach the bot not to cut URL links, but no change to the prompt has led to the desired result.

I hope that this is something that Genesys will develop further.

Kind regards,

Heikki

Hi Rajneesh,

Thanks for reaching out on our thread! 

Top questions. If you haven't enabled AI Token Experience yet, then I believe the two month offer is still applicable for any of the AI functionality. It's best to discuss this with your CSM or Partner (if you have one), to ensure you make the most out of it. 

You are right, the Virtual Agent functionality consumes one token every two interactions that go through the flow. Right now there is no additional token consumption for those that contain Guided Flows. After the 28th it will cost one token per interaction that goes through a Virtual Agent Flow that has a Guided Flow(s). This means if they are are not used at all, or if many Guided Flows are used in the one Virtual Agent flow, it will consume one token per single interaction. 

AI Studio use doesn't have any cost right now but after the 28th each new Guide created will consume one token. Hence why we're encouraging everyone to build many now, while it is free. 

Hope this helps,

Becky 

Hi,

For the 28th August timeline, is the additional cost to use AI Guides feature or how many you are create?

In recent webinar they mentioned 2 session/token pricing on usage till 28 August and later would move to 1 session / token.
We currently havent purchased AI token (which now shows 2 months trial when trying to purchase). Will this trial AppFoundry trial affect the AI guides trial?

Can you confirm any additional requirements outside license and AI tokens needed for start using AI Studio?

Kind regards
Rajneesh

Hi @Edgard Cuadra,

Great question. From my personal perspective I believe we're trying to be reserved with the release of AI-based functionality as it has the potential to cause more issues that it resolves, if not handled well. You'll likely see guardrails, like this type of limit, impossed when functionality is first released and then see them slowly being removed as performance is assured. 

Some of these limits are not fixed. If you have a good business case to request a higher limit, please open a case with our Product Support team to request an increase. If there is a good reason they cannot increase it, that will be shared. Just be wary of the potential impact to experience.

HI, 
Can you explain the reason to limiting the number of variables that can be set per guide to only 5? 

The maximum number of variables that you can have in a single guide is five.
Variables in guides - Genesys Cloud Resource Center