Is there a way to set up a Slack channel as a "Chatbot" so I can communicate with my Relevance Assistant? I can post in Slack to trigger an agent via Zapier. But I want my agent to report back to Slack, too. I would like to have a two-way conversation with my assistant agent in Slack.
i dont get it... your example is no different that using chatgpt directly. is there anything relevance can do what is not already builtin into chatgpt ?
is there any way to call another api authenticating using OAuth2? my ATS presents an API I would like to build the ability to be able to access data from the ATS Api from one of the tools and then pass it to a LLM. I see how it can be done with discord whic give you like a key and thne you pass that but authenticating with oauth is different...any ideas?
I'm having a difficult time understanding the rationale for interface or how everything relates. Usually UI design leans on the nearest mental models and metaphors to come to a layout... Maybe it'll click. In any case, I guess I'll have to keep using it to get better. I much appreciate the power, and the UI sure seems to surface a lot of affordances. Thanks!
How do you access and where do you create those core LLM instruction settings? Can't find it anywhere. I don't have this section in tools in agent settings. Thanks
Hey there! Can you link cloning your workflow? It would be nice to change out a few items and not having to start from scratch when creating something similar.
Unsure! One thing our platform does poorly is communicate credit cost. You mostly just have to check your credits before/after. We are working on this!
Hi Dan, does it mean the 'Manager' (where the Escalation goes to) needs to have a Relevance account? What if the Managers are only reachable via WhatsApp?
Yes, right now to respond to an escalation you have to log into Relevance and reply via our UI. In the future we will allow for replies to come directly from the email/Slack/future integration we escalate to.
@@danielpalmer Escalation would work best in client scenario which their staffs (with no access to relevance) could respond in real-time. WhatsApp and telegram would work best.
This is great but only partially practical. When there is no answer, the manager must manually input it before the AI Agent passes it to the user, and that may take time. AI Agent should ask for an email from the user highlighting that as soon as manager responds you will receive an email. Thank you for your patience etc. The user should i.e. receive an email with the answer after the manager responds.
I wonder if that can be included in the lead up to the hand off of the Q? Once it realises that it has to escalate it does [steps]? I agree with what you say though.
@@wnh79 Yep, great point both of you! This is right - you can definitely get it to do this by asking it to send an email before escalating to the manager :)
I would definitely like more information on this. I have just started using Relevance AI and we already have built out a beautiful sass product with Vapi AI. I really want to know more information on how these two can integrate and I know a lot more people who use Vapi AI as well and would love the information. I already have ideas and I think they are possible with Relevance AI.
Thanks for the tutorial! I will sit down and rebuild for our company. Also: How can I build an agent that has the trigger getting a database of email lists and respective emails, to then go and start sending these emails to the recipients?
The idea of using AI agents as an additional set of hands is fantastic. This could completely change the way teams work. Anticipating Relevance AI's assistance in the development of these agents and their incorporation into current workflows. Fantastic post! #SmythOS #AI #automation #AIagents
I'm a huge fan of the platform, unfortunately we aren't getting the service that SafetyCulture has received. We are unable to import the OpenAI library into Relevance AI. We have reached out to support and have had no feedback.
Hey guys, i've been really struggling to import relevance AI onto voiceflow going through agentive. When I import my agentive agent onto voiceflow everything works, but as soon as there's a relevance AI tool integrated, it gives me "Error getting Agent Status" message Would really appreciate some guidance
Tried it. The problem is that the video makes it looks easy but in fact it is quite hard to make it work without your support team which is impossible to find. They are always busy which I can understand their attention is for bigger companies etc. wish you could give us real detail docs on how to make the tools work.
Hi Warren - thanks for watching! We have a lot more resources on the way. If you have concepts or questions you'd really like us to address, we are checking our Discord where you can post suggestions.