Reports and accuracy
See how TraceX is performing for your organization, investigation volume, root causes confirmed, fix outcomes and repeat issues, and measure answer accuracy on your own systems.
Before a team relies on an AI engineer, it wants to know how often it is right and how much time it saves. Reports answers that with your own numbers, not a benchmark.
What Reports shows
Section titled “What Reports shows”| Section | What it tells you |
|---|---|
| This week | Investigations completed, how many found a root cause, and the median run time. |
| Volume | Investigations over the last 14 days, completed and halted. |
| Intake | Where issues arrive from: dashboard, Slack, email, Jira and so on. |
| Components | The parts of your system that come up most often. |
| Confirmed root causes | How many root causes your team confirmed in reviews on finished investigations. |
| Fix outcomes | What happened to fix pull requests: merged as written, merged after edits, or closed. |
| Repeat issues | Problems that keep coming back, which usually point to a deeper fix. |
ROI, insights and accuracy reports are part of the Team plan. Report history covers 30 days on Starter, 90 days on Pro and one year on Team (Plans).
Make the numbers meaningful
Section titled “Make the numbers meaningful”Reports are only as good as the feedback behind them. Ask your team to confirm or correct the root cause at the bottom of each finished investigation. It takes a few seconds, it feeds these numbers, and corrections teach TraceX for next time.
Answer accuracy
Section titled “Answer accuracy”Under Settings → Answer accuracy, admins can run TraceX against questions with known answers, such as “how many orders did we take on 1 March?”, to check its accuracy on your systems over time. Run it after connecting a new source or changing the AI model to see whether answers got better.