Virtual Agent

Hello Support community team; I escalated ticket #TS2221540. But this ticket is only Trouble shooting. From Support agent, my request should ask to this community.Because this is ZVA’s development or make call center by using ZVA’s voice agent. I think the following 5 step's Scenario.(Now Widget menu is Blank) 1. Customer call test help desk#(050-xxxx-xxxx).2.If customer wants to Japanese, Push 1.If customer wants to in English , Push 2.3.Hearing customer situation.4.If need, ZVA's Voice agent ask to customer for catch up detail.5.If ZVA's agent understands, he or she suggests accurate solution.6.If above's suggestion's does not fixed, Voice agent ask to customer name and mobile number.7.Can we receive from step 3 to 6 conversation's story by e-mail?8. Form improvement of speech accuracy, I'd like to manage knowledge base, add too, and Intent function.(But I'm not sure their steps)

Hey Folks!Depressingly this video is already out of date as 3.0 very recently launched, but hopefully it will still help you familiarise yourself with ZVA build & moving away from the classic chatbot 1.0 concepts.https://youtu.be/3PjVg2hfq6k

If you are asking a series of questions to collect information but if the customer’s answer then determines the next question (an “if-then-else” prompt) this is the best practice with ZVA Agents: 1. Turn on AIC 3.02. Make an instruction skill 3. Limit “if-then-else” prompts to five scenearios using an instruction skill. If you use  six or more “if-then-else” statements in your prompt, then the experience will degrade because the context of the conversation is not best matched for one agentic conversation. If you have more than five “if-then-else” statements in a single prompt, use these questions to continue following best practices: » Why do we need these questions? » Can we include the intent of these questions inside of questions #1-5? » Will my customer be fatigued at this point and give false or incomplete answers in order to “just get it done?” » Have you prepared the caller before s

 ⚫️ Process to test AIC 3.0 and transition your current agent to AIC 3.0:Duplicate your existing agent Ensure the ZVA 3.0 toggle is enabled on the cloned agent Run test scenarios in preview to confirm expected behavior (Optional) Once validated, go to your production agent and scroll down to the bottom of the page to find the traffic splitting setting (Optional) Start by routing 10% of traffic to the cloned agent with ZVA 3.0 enabled, 90% of traffic to your existing agent. (Optional) Monitor containment rate, resolution rate, and surveys to identify any adjustments needed in the ZVA 3.0 agent. Make adjustments and publish the AIC 3.0 agent. Gradually increase the traffic routing until you are ready to enable ZVA 3.0 on your production agent. Once you are at 100% of traffic on the AIC 3.0 agent, then you can go back to the AIC 2.0 original version, turn on AIC 3.0 and 100% of traffic

I’ve been playing with ZCA quite a lot recently & just added the latest video to the #apieceofCX library with a look at ZVA Analytics.Very impressive & super accessible.Watch on YouTube here!

 Check out the features from January that we recommend you consider for your ZVA management practice! All features listed here: https://support.zoom.com/hc/en/article?id=zm_kb&sysparm_article=KB0067812

Hi everyone. I'm trying to setup the integration between AVZ and Manage Engine ServiceDesk Plus.Based on the KB https://support.zoom.com/hc/en/article?id=zm_kb&sysparm_article=KB0057946 I insert "itdesk" and select my region (EU). I'm redirected to the auth page on Zoho but after the authentication process I land on an error page that says:  Invalid ClientClient ID passed does not exist Any idea?Is the ServiceDeskPlus supported or do I miss something?"itdesk" refers to other ManageEngine products? RegardsLuca  

Hello, I need your helps. I got 050 phone number by eKYC and paid “Charge Name:Zoom ContactCenter Monthly Usage -”. But I could’t find voice agent menu(I showed chat agent and classical chat bot menu only).I’d like to make the simple IVR(Japanese or English and Loss the urgent matter(mobile , PC) or Admission Application) using Zoom Virtual Agent simply.Would you mind sharing the step by step the flow guide immediately?Also I’m very happy to the guide with Japanese directly. I appreciate your dedicated support for Japan in advance, Hiro.  

I would like to be able to filter matched queries that I have already coached and trained. I would expect coached queries to drop from the list of matched queries or there should be a filter to hide "coached queries". Do they never disappear? If not, how is that scalable? If they should be dropping from the list, what step am I missing? My steps are:1. Navigate to Intent Management.2. Scroll to an Intent group.3. Select an intent within the group.4. Click on the Coach tab.5. Click the check box next to queries to be coached.6. Add to training phrases.7. Train.8. Publish. The use case for why it is problematic is sorting through multiple identical or similar queries to find the queries that I have not coached. Below is a real-world example of one of my intents:1. I have an intent for Financial Input.2. I have 195 queries for "Financial review," 4 queries for "Financial input," and 1 query for "Ca

We have a complex contact center flow that is working well but we cannot stop the transfer tool from saying its inner chain of thought inspite of various guardrails.Any ideas? None of our prompts provide instructions on reading content back to the caller.eg. AI Agent Hello. How can I help you today?Consumer How can I set up a Saturday morning detention?AI Agent Okay. Let me transfer you to one of our support agents who can help with this. Can you please tell me any more details we may need to know before I transfer you?Consumer Uh, our period structure is just a Monday to Friday.AI Agent Right. Thank you for that detail. Is there anything else you'd like to add before I transfer you?Consumer No.AI Agent Setting up a Saturday morning detention. The period structure is currently Monday to Friday

Make sure that you are optimizing your agent for the job they were “hired” for. Do not prompt-bomb the guidance and write a Da Vinci Code length guidance prompt. It’s hard for a human to read and hard for AI to understand. So, what is the ideal length of your guidance? First, factor in that ZVA is not only digesting your guidance prompt but also, Length of the conversation - which we are optimizing using summary. Number of skills and their complexity. How tool responses are being returned. How guidance is written as well (the architecture and formatting matter).  Best Practice:Target for one agent’s guidance prompt: 2,000–4,000 characters(~500–1,000 tokens)Why this range works bestEnough room for:role definitiongoalstoneguardr

I am developing a custom solution using Zoom Virtual Agent (ZVA) where customers provide image inputs during chat interactions. However, ZVA does not support image uploads out of the box, as confirmed in my previous case with you. In the CX Analytics tab > Logs > Transcripts, I can see the chat history, including attached images , which appear in the UI. To work around this, I want to programmatically fetch the transcript and the image attachment using APIs, as text responses can be passed via variables, but images require API access for processing (e.g., in AWS Lambda). I have explored the following APIs with a Server-to-Server OAuth app (scopes include contact_center:read:engagement:admin and contact_center:read:inbox_messages:admin): GET /contact_center/engagements/{engagementId}/events Response: {"code": 7001, "message": "Only voice and video channels are currently supported."} Limitation: Does not support messaging channels. GET /contact_center/messaging/messages?engagem

hi Zoomies.  I didn't see a topic area for AIEA, so i am posting here in ZVA.  I am setting up a simple chat flow and queue as a customer entry point for general inquires.  The queue will have the Expert Assist enabled (KB and NBA).   Since you can only have one AIEA instance assigned to a queue.   AND, since each AIEA can only have one Intent Group. (I think) Then, if your queue is for general inquiries,  won't your Intent Group have to be very broad (and possibly large). I'm hoping someone can tell me I'm missing something.   In a perfect world, it would be nice if I could assign multiple Intent Groups. 

Hello, My company has signed up to Zoom Contact and Zoom Phone Enterprise. I have a good working knowledge of AI voice agents having worked on POC agents in Ragflow, N8N and Elevenlabs. It's early days so I am happy to park this to one side as we are working with a professional partner.What I am struggling with is the change in behaviour between querying a knowledge base, querying a virtual agent with a minimal prompt and querying a  deployed virtual agent in full screen url campaign deployment. Given I have spent much of the last year, preparing, cleaning and building RAG agent  workflows for text and speech, I am trying to figure out why the same knowledgebase would retrieve different matched content from the same knowledge base using a vanilla prompt.Are there video tutorials and documentation that go through end to end example projects to give blueprints and examples for start to finish workflows, as is often the case with N8N

When the Zoom Virtual agent submits a response, it plays a tone at maximum volume, which hurts my ears. The tone does not respect the Windows volume control. Environment: Dell Latitude 5400 laptop, Windows 11 Pro, current Zoom version.

If your knowledge base sync is stalling or creating an error due to size, please read this article. https://support.zoom.com/hc/en/article?id=zm_kb&sysparm_article=KB0058272If you're KB is under the size limit but still stalling when syncing, please submit a support ticket under the product = zoom virtual agent. On Monday mornings, the KB crawlers are slowest because clients like to update their KB's then. Try an off-rush hour time like Monday night or Tuesdays for a faster weekly sync. 

We are preparing for the transition to Salesforce Enhanced Messaging from Live Chat. As a company with support agents around the clock and around the world, we rely heavily on a script to check agent availability in Live Chat. As Salesforce will not be providing a similar solution for Enhanced Messaging, I'm curious how other companies are addressing this critical workflow. Suggestions?

I have a zoom hearing in court coming up. How do I upload exhibits to show to the judge?

I set up our meeting to include the "registration required" so that I could track the log-in and log-out of each participant. Everyone registered; however, there is no report available. I've used this feature on other Zoom accounts and it work fine. What exact setting do I need to make sure it enabled so that I can track the attendance of registrants/participants?

For one virtual agent, best practice is is 5 to 6 skills maximum. Within each skill, you can have 4 to 6 tools listed.  Ensure your skills are distinct (description) and tools have good description as well.

If you are using ZVA Chat 2.0 agents to escalte to a video call, you will need to whitelist the domains below:ZCC USA-Based CCIZP https://us01cci.zoom.us/", CCIZPAPI

We recently updated our IP addresses for web crawlers. If you are seeing an error in your knowledge base sync, please ask your IT team to whitelist/allow these IP addresses to crawl your website.  44.199.223.208 44.198.248.168 44.199.239.85 3.12.25.1 3.12.48.211 3.12.140.78Additionally, depending on your website architecture, you will want to check that your website allows for Zoom.us to crawl and index in your /robots.txt files. Also verify that your website's headers have "<meta name="robots" content="index,follow"> to ensure Zoom has access to crawl the sites you would like. You can limit the pages and content that you want Zoom to have access to by adjusting your robots.txt file. Try enabling the "javascript support" feature in the know

Hello, I'm working with Zoom Virtual Agent and wanted to clarify two things: User-Specific Knowledge:Is it possible to assign or tie specific knowledge to an individual end user, so that when they ask questions during a session, the bot references only (or prioritizes) that knowledge?Our intended workflow is:We receive the event that a chat has started.We look up relevant data for that user in our database.We want to pass that data to the bot as knowledge, so it can use it in responses during the same session.Dynamic Knowledge via API (Mid-Session):If we add new knowledge via the Knowledge API during an active chat session, is the virtual agent able to immediately access and use that information in the same ongoing conversation?

If you have successfully created a campaign and see the "button" for Zoom Virtual Agent on your website, but it's not loading, then it's likely the SDK has not been "whitelisted by your website's CSP.Some websites have a Content Security Policy (CSP) that only authorizes "whitelisted" or "allowed" endpoints to access their website. Each CSP is unique, so please get in touch with the person who manages the website you want ZVA to work on (or across) and ask them to update the CSP settings to match the recommendations in this article. CSP SETTINGS FOR ZVA SDK: https://developers.zoom.us/docs/virtual-agent/web/get-started/Send this link to your web developer, and they will understand which protocol you use and implement the CSP headers for ZVA accordingly. Additionally, please note that Zoom's SDK options are committed to sec