AI & Bots

Most teams migrate to Freshdesk, turn on Freddy, and wonder why accuracy is low. The answer is almost always the same: the data went in the wrong order.Why Freddy fails after most migrationsFreddy AI draws on two sources to do its job: your Knowledge Base and your closed-ticket history. Both matter, but the sequence in which they arrive on the platform matters just as much as the content itself.If you migrate tickets before your KB articles are in place, Freddy starts indexing ticket content that references articles it cannot yet retrieve. The AI routes intent correctly but generates responses from a blank KB — or worse, hallucinates content that isn't there.If you migrate unresolved or low-rated tickets, the problem runs deeper. Freddy's auto-triage and article suggestion features l

Current flow:Customer send mail to support mail New ticket with status open is created Freddy replies with solution Customer does not react to Freddy: Ticket stays open → but what if thats already the “best answer”? Customer closes the ticket → everything is good Customer says the answer is not good My question: Wouldn’t it be better to set the ticket always to status resolved when Freddy gives a reply to the customer and only if the customer clicks “not good” or replies we open the ticket?How should our agents handle tickets that are open but already contain the right answer from Freddy?Are there any best practices?Thank you so much! 

Hello Freshworks Community!Our next learning session on Designing and Deploying Agentic Workflows is about to begin in just a couple of hours!Here’s your direct link to join the session 👉https://freshworks.zoom.us/s/91882472858?pwd=tvVj0iSFD4p6O9r2Qf5ZA5Hp2XhP28.1 We’re super excited to see you all there! Get ready for a deep dive into Vertical AI Agents, agentic workflow design, and practical tips to help you go live with Freddy AI faster.See you soon! 

Hello Freshworks Community !! We just wrapped up Refresh North America 2025, where we unveiled powerful CX innovations designed to help your teams support customers better, resolve faster, and scale with confidence. Here’s what’s new: Freshdesk Omni with the Command Center The Command Center brings together conversations from across channels, context, and intelligence into one centralized workspace. Agents can personalize interactions and resolve issues faster without switching tabs or juggling multiple tools. Admins gain full control over workflows and performance, with flexible options to configure routing, SLAs, and channel-wise availability. Agentic AI is embedded throughout, delivering out-of-the-box intelligence fro

I recently faced the same challenge and started exploring different Dubbing AI tool. After testing several platforms, I found Dubbing AI to be one of the most efficient and accurate solutions out there. It uses advanced AI voice cloning, translation, and lip-syncing to make your videos sound natural in over 150 languages.What I like most is that it maintains your original voice tone and style, so the result feels completely authentic. For anyone creating educational, business, or entertainment videos, Dubbing AI makes global content creation super easy and affordable.

Hello,  Our team currently use the ‘FreshService Chatbox’ and we wanted to see if it was possible to change the background color and also make the button “IT Support Chat” slightly bigger.  Could someone please advise how I can proceed? I searched in Admin → Setting → FreshChat but wasn’t able to find a way. Not sure if I need to code it in ‘Service Desk Rebranding’ page but a confirmation or suggestion will be great.   

I've enabled the Thank You Detector and made the necessary changes to the ticket reopen rule, but it still isn't working as expected. It's not even detecting a simple "Thanks" message. Can someone help?

Hey everyone,Today’s session — Community Botcast: Freddy AI Agent — is about to begin, and we’re excited to have you join us!If you haven’t received the link yet, you can join directly here:Join the sessionIf you can’t make it live, don’t worry — we’ll share the recording in this thread afterward.See you soon! 

Hello Folks!Here’s a quick look at what’s new with Freddy AI for Customer Service this month. These updates are aimed at improving both agent productivity and admin control—plus, a few exciting features are now in public beta! What’s new with Freddy AI for customer service this month? There’s a lot of new updates and enhancements that we’re rolling out to customers using our customer service products with Freddy AI! Here’s a quick summary for everyone here. Reply Suggester is now in Public Beta:  Agents now have access to AI-powered reply suggestions (generated using solution articles) to respond to incoming customer queries, resolving tickets faster and more accurately.Learn more |  

Heyy Folks!Struggling to decide between a Chatbot and an Agent Assistance Bot for your customer service? Join us for a 60-minute virtual session where our Subject Matter Expert, Neil Dutta will break down the differences, key use cases, and when to use each of it for maximum impact. Here’s the registration link -  Direct Joining link if you haven’t received the Zoom invite. 👇 https://freshworks.zoom.us/webinar/register/WN_SrBev2SdSqeqoZ8MyQ_JhA Please ping me on this thread in case you face any issues with the joining link. Happy to assist.

I’m looking to better understand how many people use the Auto Email feature in Freshdesk.In it’s current form it’s an On/Off option and you are unable to limit what Groups or other ticket fields it can be enabled for. I’ve raised on the Ideas Forum () however I’d like to know if any other companies are using the feature.I don’t want to enable it on all tickets, and instead limit it to doing First Response on non SLA contracted questions/account problems, rather than big P1 total loss of service tickets.Thoughts? 

Both email bot and Freddy AI are missing from my admin page.What is the solution to get them to display? David

Hello, I am quite new to bot building so I hope someone can help.I have created a flow where a reference number is collected. This has the following regex validation applied:(700000[0-9]|70000[1-9][0-9]|7000[1-9][0-9]{2}|700[1-9][0-9]{3}|70[1-9][0-9]{4}|7[1-9][0-9]{5}|[89][0-9]{6})I have then created a condition where if the response is valid you to go dialogue 12.2, else dialogue 12.4.Where a valid reference number is entered this works as intended, however if something else is entered this is not going to 12.4, instead it is just asking for the reference again (12.1) leaving you stuck in a cycle.Is there something I have done wrong? How can a you get out of this cycle without entering a valid reference?