Using Recommended Resolutions

Recommended resolutions surface similar past tickets from your team’s history that may help you resolve the current ticket. Think of it as having a searchable memory of every issue your team has handled before — except the AI does the searching for you.

What Are Recommendations?

When you open a ticket, Flexivity AI automatically compares it against your team’s resolved tickets. If it finds past tickets that dealt with similar issues, it presents them as recommendations. Each recommendation includes:

Ticket subject and summary — a brief description of the past ticket so you can quickly judge its relevance.
The resolution — how the past ticket was resolved, giving you a starting point for the current issue.
A similarity score — a percentage indicating how closely the past ticket matches the current one.
Up to five related tickets — recommendations include up to five past tickets with links, sorted by relevance.

Where Do They Appear?

Recommendations appear in the Flexivity AI panel on the ticket detail page, alongside the AI summary.

If no similar tickets are found, the recommendations section will simply indicate that no matches were found. This is normal for unique or new types of issues.

How to Interpret Similarity Scores

Each recommendation comes with a similarity score expressed as a percentage:

80% and above — Highly relevant. The past ticket is very similar to the current one and is likely worth reviewing.
60% to 79% — Moderately relevant. The past ticket covers a related topic but may not be an exact match. Still worth a quick look.
Below 60% — Lower relevance. The past ticket has some overlap but may address a different root cause. Use your judgment.

Tip: Similarity scores are based on the content and context of the tickets, not just keyword matching. A recommendation with a high score may use different terminology but describe the same underlying issue.

Note: If you see lower similarity scores across the board, it may simply mean that your helpdesk has a smaller ticket history for comparison. As your team resolves more tickets, the knowledge base grows and scores will improve.

How to Apply a Recommendation

When you find a recommendation that fits the current ticket:

Review the suggested resolution. Read through the past ticket’s resolution to understand the steps that were taken.
Adapt it to the current situation. Past resolutions are a starting point, not a script. The current ticket may have slightly different circumstances, so adjust the response as needed.
Use the apply action if available. Depending on your configuration, you may see an option to apply the recommendation directly, which pre-fills a response that you can edit before sending.

Remember, recommendations are suggestions — you always have the final say on how to respond to a ticket.

Providing Feedback on Recommendations

Like summaries, each recommendation has thumbs up and thumbs down buttons. Feedback is completely anonymous — no user information is captured or tracked.

Thumbs up if the recommendation was relevant and helpful, even if you did not use it word-for-word.
Thumbs down if the recommendation was off-target or not useful for the current ticket.

Your feedback is tracked in the analytics dashboard and informs manual prompt adjustments that improve how future recommendations are ranked and presented.

Tip: Even if you do not end up using a recommendation, giving it a thumbs up when it was relevant helps the team understand what “good” looks like for your ticket types.

Recommendations Improve Over Time

The recommendation engine is powered by your team’s own ticket history. This means:

The more tickets your team resolves, the better recommendations become. The knowledge base grows with every resolved ticket.
New teams or new issue categories may have fewer recommendations at first. This is expected — coverage improves as the system processes more data.
Your feedback informs improvements. Ratings help administrators and the Flexivity AI team identify which patterns from your ticket history are most useful, so they can make targeted prompt adjustments.

Tip: If you are not seeing many recommendations yet, that is normal for a team that recently started using Flexivity AI. As your resolved ticket count grows, you will see more and more relevant suggestions.