A guide to using Freshdesk Freddy to spot knowledge gaps from unresolved tickets

Unresolved tickets are more than just a queue-clogging headache. They're actually a goldmine of information, pointing you directly to the questions your customers have that your team (and your knowledge base) can't answer. The problem? Nobody has time to manually sift through thousands of tickets to find these "knowledge gaps." It's a soul-crushing task.

This is where AI is supposed to step in. Freshdesk’s native AI, Freddy, promises to deliver insights that can make this process easier. But how well does it actually work for spotting gaps in your support content? This guide will walk you through how to use Freshdesk's tools for ticket analysis, where you'll likely hit a wall, and how a more focused approach can automate the entire process.

Understanding the role of Freshdesk Freddy

Before we get into the weeds, let's get a clear picture of what we're working with. Freshdesk Freddy isn't a single product. It’s the brand name for the AI engine that Freshworks has woven throughout its customer service platform. The goal is to help on two fronts: making customers more self-sufficient and making agents more productive.

For what we’re talking about today, it breaks down into a few key pieces:

• Freddy AI Agent: This is the autonomous part of the AI that handles customer-facing chats and emails, providing direct, automated replies.

• Freddy AI Copilot: Think of this as the agent's sidekick. It works inside the help desk to help human agents summarize long ticket threads, draft replies, and suggest relevant knowledge base articles.

• Freddy AI Insights: This is the analytics engine. It looks at the numbers behind your support operations to show you trends, helping you make better decisions about staffing, performance, and potential issues.

Together, these parts create an AI layer within Freshdesk. Now, let’s see how you can use its reporting features to start hunting for those knowledge gaps.

How to use Freshdesk reporting to spot knowledge gaps

If you're trying to find knowledge gaps with Freshdesk's built-in tools, you won't find a magic "show me what's missing" button. The process is more of a detective job, relying on you to use its help desk reports to piece together clues from unresolved tickets.

Here’s how you can tackle it.

Analyzing ticket volume and distribution reports

Your first stop should be the analytics dashboard. With a few filters, you can start to see patterns emerge from all the noise.

• Ticket Volume Trends: Start by creating a report that tracks your ticket volume over time, but filter it to only show tickets with an "unresolved" status. If you see a sudden spike after a new feature release or a marketing campaign, that’s a huge red flag. It probably points to a new, undocumented issue that’s catching everyone by surprise.

• Ticket Distribution by Category: A more detailed approach is to look at unresolved tickets based on how they're categorized. If you use tags or custom fields, you can group them to see which topics are causing the most trouble. For example, if you notice a ton of unresolved tickets are all tagged with "annual-plan-refund," you’ve just found a major gap in your billing docs.

Using customer satisfaction (CSAT) feedback

Unresolved tickets and unhappy customers often go hand-in-hand. Your CSAT surveys are a direct line to understanding why a ticket couldn't be resolved.

Create a report that filters for tickets that are both unresolved and have a negative CSAT rating. The open-ended feedback here is pure gold. Customers will often tell you exactly what they were looking for and couldn't find. You’ll see comments like, "I was trying to figure out your international shipping policy but there was no information on it."

The limited role of Freddy AI Insights

Freddy AI Insights can speed up some of this work. It’s designed to automatically flag high-level trends, which can save you some initial digging. For instance, it might highlight that "password reset" is a recurring topic in your support queue.

But, and this is a big but, that's pretty much where its usefulness ends. Freddy might tell you what customers are asking about, but it doesn't confirm if a knowledge gap is the root cause, and it certainly doesn't help you fix it. A human still has to connect the dots, confirm the gap exists, and then go write the missing content by hand.

Key limitations of using Freshdesk Freddy

While you can technically use Freshdesk's reporting to find knowledge gaps, the process is far from smooth. Once you start digging in, you'll quickly run into some pretty big drawbacks that keep it from being a truly automated or forward-thinking solution.

A manual process

The biggest headache is that the whole process is manual. Freshdesk’s reports give you the raw data, but it’s up to a support manager or an analyst to do all the heavy lifting. You have to spend hours staring at dashboards, reading through individual ticket conversations, and trying to guess what’s missing.

This means you’re always playing catch-up. You only discover a knowledge gap after dozens, or even hundreds, of customers have already hit the same wall and had a frustrating experience. You're constantly putting out yesterday's fires instead of preventing tomorrow's.

Siloed knowledge

Freddy AI is trained almost exclusively on the data that lives inside your Freshdesk account, your help center articles and ticket history. The problem is, that's rarely where all the answers live. Critical information is often scattered in other places, like a technical guide in Confluence, a policy document in Google Docs, or a product update on a Notion page.

Since Freddy can't access these external sources, its knowledge is incomplete. It will fail to answer questions even when the information is available somewhere else in your company, which just leads to more unresolved tickets. The AI is stuck inside its own little box, unable to see the full picture.

No proactive simulation

With Freshdesk, you have to turn the AI on and let it run with live customers before you can get the data you need to find its weaknesses. There's no real way to test how Freddy will perform on your past tickets to find these gaps before it starts talking to people.

This is a risky way to work. You’re essentially launching an AI knowing it has blind spots, but you won't know what they are until real customers find them for you. It’s like sending a new agent to the front lines without knowing if they can answer even the most basic questions.

An alternative approach: eesel AI

The shortcomings of Freshdesk’s approach show why you need a tool built from the ground up to be proactive, not reactive. This is where eesel AI makes a real difference. It’s designed to automate the entire workflow of finding and closing knowledge gaps, so you can get ahead of problems instead of constantly chasing them.

Automate knowledge gap identification and content creation

Unlike Freshdesk, which just dumps data on you, eesel AI’s analytics dashboard gives you insights you can actually act on. It doesn’t just show you what your AI is doing; it pinpoints exactly what it couldn't do. It highlights trends in unanswered questions, giving you a clear, prioritized list of your biggest knowledge gaps.

Even better, eesel AI helps you close those gaps. It can automatically generate draft knowledge base articles from successful resolutions in your past tickets. If an agent has already written a perfect explanation, eesel AI can turn that into a new help center article with a single click. You end up building a knowledge base with content that’s already proven to work.

Unify all your knowledge sources instantly

While Freddy is stuck within the Freshdesk ecosystem, eesel AI connects all your scattered sources of information. In just a few minutes, you can link it to all the places your team stores knowledge.

Whether your answers are in Google Docs, Confluence, Notion, or another tool, eesel AI can learn from it. It also trains on your past ticket history from help desks like Freshdesk, giving it a complete picture of your business from day one. This means fewer unresolved tickets because the AI has access to all the answers, no matter where they are.

Test with confidence using powerful simulation

Perhaps the biggest difference is eesel AI's powerful simulation mode. Instead of "going live and hoping for the best," you can run the AI on thousands of your historical tickets in a completely safe, sandboxed environment.

This lets you see exactly how the AI would have responded to real customer queries. You get an accurate forecast of your potential resolution rate and, most importantly, you can proactively identify knowledge gaps before a single customer is affected. The simulation report shows you which types of tickets would have been escalated, giving you a clear roadmap of what content you need to add before you even think about turning the AI on.

Freshdesk Freddy AI pricing

To get the full picture, it helps to look at Freshdesk's pricing for its AI features. It's split into two main add-ons, which can make costs a bit complicated and hard to predict.

• Freddy AI Copilot: This is the agent-assist tool, and it’s priced per agent. It starts at $29 per agent, per month when you pay annually.

• Freddy AI Agent: This is the autonomous bot, and it’s priced based on "sessions." You get some free sessions to start, but after that, you have to buy packs, which cost $100 per 1,000 sessions.

A "session" is a single email reply or all bot interactions in one chat over 24 hours. This means a single email ticket that needs a few back-and-forths with the AI could burn through multiple sessions, causing your costs to jump unexpectedly. This session-based model can make budgeting a real challenge, which is a stark contrast to the transparent pricing of alternatives like eesel AI.

Final thoughts on using Freshdesk Freddy

At the end of the day, an unresolved ticket is a direct signal that your support documentation is failing your customers. While you can use Freshdesk Freddy to spot knowledge gaps from unresolved tickets, the process is overwhelmingly manual, reactive, and limited to the data inside the Freshdesk platform.

Freshdesk’s reporting can give you clues, but it still takes a ton of human effort to turn that data into real improvements. A modern support team needs a better way.

Tools like eesel AI are built to automate this entire workflow. By connecting all your scattered knowledge, running powerful simulations on your historical data, and automatically flagging gaps, you can shift from reacting to problems to proactively solving them.

Get ahead of your unresolved tickets

Stop wasting hours digging through reports. Let AI do the heavy lifting of finding and closing your knowledge gaps.

eesel AI plugs directly into your existing help desk like Freshdesk in minutes, so there's no need to overhaul your current setup. You can run a simulation on your own historical tickets instantly and see exactly where you can improve.

Sign up for free to see what knowledge gaps you can spot today.