TABLE OF CONTENTS
- Freddy AI Insights for Freshservice
- Root Cause Analysis (RCA)
- Conversational Insights (Freshservice)
- Conversational Insights - Service limits
- General Availability (GA)
Freddy AI Insights for Freshservice
1. How do I enable Freddy AI Insights?
Follow the steps below to enable Freddy AI Insights:
Enable for an account: Admin Settings > Freddy AI > Toggle ON ‘Freddy AI Insights’.
Enable for an users: User Management > Agents > Select User > Permissions > Edit 'Groups and Roles' > Add ‘Freddy Insights’ Permissions.
2. Who is Freddy AI Insights for?
Freddy AI Insights is essentially for supporting supervisors and decision-makers. With AI-powered conversational inquiries and auto-generated insights, they can gain visibility into IT performance.
3. What languages are Freddy AI Insights available in?
Freddy AI Insights is available only in English at the moment. The Multilingual Insights feature is in our product roadmap.
4. Can we export reports from Freddy AI Insights?
This is not supported at the moment.
5. What are the types of insights available?
6. Why am I unable to find any insights?
The following conditions need to be met for insights generation:
Once the Freddy AI Insights module is enabled (i.e. the toggle is turned on), your service desk insights will be available after 24 hours.
The eligible group of the account(s) should have a minimum of 10 tickets for generating insights in the last 12 weeks.
Only tickets assigned to groups will be considered for insights generation.
Insights are generated when the algorithm identifies anomalies or trends in the dataset. There can be scenarios where there are enough data points but little variation, and hence, no insights will be generated.
7. What duplication checks are done for various types of insights generated?
Majority: Same category value within ±5% contribution across a time period.
Outliers: Identified on the same date range.
Trend Change (Spike & Fall): Identified on the same date range.
Longest Increase/Decrease: Identified on the same date range.
Recent Change: No duplication check.
Overall Trend: ±5% magnitude in the same direction.
8. At what aggregation level does Freddy AI generate insights?
Freddy AI Insights are generated at an Agent Group level at the moment.
9. For which modules does Freddy AI generate insights?
At the moment, AI Insights are generated only for the ‘Tickets’ module.
10. How frequently will Freddy AI generate insights (Proactive Insights and RCA) for my account?
Note: Root Cause Analysis (RCA) is pre-computed, and data points may vary with the current number (based on the above refresh frequency).
11. Can I customize the insights that Freddy AI generates in my account?
Yes, click here for more details.
12. Are the AI Insight report formats customizable?
At the moment, this is not available, but it is on our roadmap.
13. Will AI Insights consider data migrated from other systems for Proactive Insights and RCA?
Yes, data migration from other systems is considered when generating AI Insights. However, we need to ensure two things:
Tickets should not all have the same creation date.
The agent groups referenced in the tickets must exist in our system.
14. Are notifications available on email/Slack/Microsoft Teams when an insight is generated?
Not at the moment, but this is on our roadmap.
15. Will Freddy AI Insights be available to customers in all regions?
Yes.
Root Cause Analysis (RCA)
16. Explain the frequently used terms for Root Cause Analysis (RCA).
Tree chart: The visual representation for showcasing the interlinkage of causes.
Nodes: Boxes in the tree chart showcase data, representing the underlying causes.
Critical Path: The leading/most significant underlying root cause.
Summary: A consolidated summary of all the data in different nodes of the tree chart, in simple natural language.
17. Where can I view RCA?
RCA can be accessed from the Freddy AI module in the product.
18. Is RCA applicable to all types of insights?
At the moment, RCA is generated for only Proactive Insights, which display a trend that is not beneficial or desirable for the service desk (can also be called negative insights).
19. What are negative insights?
Insights where the concerned metric has moved in the undesired direction. They are shown in Red, Amber and Yellow colours as per criticality.
Red Colour - High Criticality
Amber Colour - Medium Criticality
Yellow Colour - Low Criticality
20. What are positive insights?
Insights where the concerned metric has moved in the desired direction. These are shown in green colour. RCA is not available for positive insights.
21. What timeframe is used to create the RCA for an insight?
This depends on the type of insight as follows:
22. What types of fields and metrics is RCA applicable to?
23. What do the terms “unknown” and “others” mean in RCA?
Unknown: This refers to the empty value in any field. For example, if no agent is assigned to the tickets in question, it will show as ‘Agent Name: Unknown’ in RCA.
Others: This refers to the multiple values in the data of any field without any significant portion from a particular value. For example, out of 100 tickets, 50 were handled by Agent A, 30 by Agent B, 10 by Agent C and one each by 10 other agents. Then the RCA drill down will roll up these 10 tickets and call them out as ‘Agent Name: Others’.
24. What tickets are part of the Leading Causes underlying data?
All the tickets present in the leaf node (last node) of the critical path of the tree map.
25. What tickets are part of the View all causes underlying data?
All the tickets for that relevant week or month for which the insight is generated.
26. Can I add or edit any particular field in the treemap for RCA?
No, at the moment only the fields identified by the AI model are part of the treemap.
27. Does RCA support multiple languages?
No, at the moment it does not support multiple languages. The Multilingual Insights feature is in our product roadmap.
28. How does the RCA algorithm work?
Detects insights (anomalies, spikes, or trends) in a specific time frame.
Analyzes all underlying data and attributes influencing the insight.
Ranks top or most impactful contributors (numerical or categorical).
Iteratively adds the next most important factors from the remaining attributes until no further reasonable splits are possible.
Identifies the critical path, highlighting the most significant or highest contributing set of attributes (that is, root cause) behind the insight.
Conversational Insights (Freshservice)
29. What is Conversational Insights?
Conversational Insights lets you ask questions in plain English and instantly get visual reports or charts from your Freshservice data. You don’t need to create reports manually—it’s as simple as chatting with Freddy.
30. When is Conversational Insights available?
Conversational Insights is Generally Available (GA) for all eligible accounts as of May 14, 2026.
31. Who can access Conversational Insights?
The feature is available to customers on the Freshservice Enterprise Plan who have Freddy AI Insights enabled.
Pre-enabled Accounts: Customers who enabled Freddy AI Insights prior to August 25, 2025, will receive automatic access to the Conversational Insights on May 14, 2026.
- New Activations: For eligible accounts enabling Freddy AI Insights after May 14, 2026, the Conversational Insights feature will be provisioned and accessible within 24 hours of enablement.
32. How do I enable it for my account?
Go to Admin → Freddy AI → Freddy AI Insights and turn on Proactive Insights.
33. How do I give agents access?
Admins can go to Admin Settings → User Management → Edit Permissions and add the Freddy Insights permission for selected agents.
34. Where can I find Conversational Insights in Freshservice?
Click the Freddy icon in the Freshservice menu and open the Conversational Insights tab to start asking questions.
35. How do I use it?
Simply type a question like:
“Show unresolved tickets by priority.”
Freddy will generate the right chart for you. You can refine your query with follow-ups like:
“Only for the last 30 days.”
36. Can I add these charts to my Analytics reports?
Yes. Click Add to Report on any chart you want to include in your report.
37. Does it remember my previous questions?
Yes. Conversational Insights keeps your chat history as context for follow-up questions. To start fresh, click Clear Conversation.
38. Can I see how the chart was created?
Yes. Click the info (i) icon on the chart to view the exact query Freddy used.
39. What kind of questions work best?
Ask simple, direct questions about what, when, or who—for example, counts, trends, or comparisons. Avoid long or complex questions; instead, break them into smaller follow-ups.
40. Does it explain “why” something changed?
Not yet. Conversational Insights focuses on descriptive charts. For deeper analysis on why something changed, use Root Cause Analysis (RCA) in Freddy AI Insights.
41. Which languages does it support?
Currently, it supports English only.
42. Can it use data outside Freshservice?
No. It only analyzes data within Freshservice.
43. How accurate are the results?
Accuracy depends on the quality and completeness of your Freshservice data. Freddy may occasionally make mistakes, so always review results before taking action.
44. Any tips for getting better answers?
Ask one question at a time
Use follow-ups to refine your results
Click the info (i) icon to verify the query if a chart looks off
45. What are some example prompts I can try?
“Top agent by FCR this month”
“Compare resolution time for the last two quarters”
“Tickets by source in the past 7 days”
46. What if I enabled Freddy AI Insights after May 24, 2026?
You’ll get access to Conversational Insights within 24 hours of enabling Freddy AI Insights.
Conversational Insights - Service limits
47. What is the query limit for Conversational Insights?
Conversational Insights has a monthly service limit of 100 queries per account.
48. What is a service limit?
A service limit is a threshold that is not technically enforced. It provides a benchmark to help you evaluate the service while maintaining full access to all features.
49. What happens if the account exceeds the 100-query limit?
If you exceed the limit, the following conditions apply:
- No service interruptions: Your team can continue to run queries beyond the 100-query threshold without being blocked or throttled.
- Full exploration: The experience remains open and unrestricted to allow for comprehensive testing of the service's capabilities.
- Advanced notice: We are committed to transparency regarding your usage. We will send a formal notification one month prior to enforcing these limits.
- No current in-product indicators: During this promotional phase, you will not see usage counters or "limit reached" messages within the user interface.
50. How do I track my query usage?
Because the limit is not enforced, there are currently no in-product indicators or counters in the UI. You can continue to use the service without monitoring query counts.
51. Why is there a limit if it is not enforced?
The service limit helps you understand the intended service capacity while allowing you to explore the tool without friction during the early stages of adoption.
General Availability (GA)
52. When is Freddy AI Insights going live in GA?
Freddy AI Insights went GA on June 11, 2025. The specific Conversational Insights capability within that suite will go GA on May 14, 2026.
53. Which plans will have access to Freddy AI Insights at GA?
Freddy AI Insights is available only to customers on the Freshservice Enterprise plan.
54. I am on the Freshservice Pro plan and have been testing Freddy AI Insights in public beta. Is there any action for me?
Please upgrade to the Freshservice Enterprise plan to continue using Freddy AI Insights.
Related links:
- Freddy AI Insights for Freshservice
- Freddy AI Insights Overview
- Root Cause Analysis (RCA) in Freddy AI Insights
- Create Customized Insights in Freddy AI
- Introducing Conversational Insights
- Freddy AI Insights | Freddy AI Trust: Frequently Asked Questions (FAQs)