A New Tool for the D365 F&O Toolkit
This week on a customer engagement, we tested a new Copilot agent from FastTrack: the Trace Parser Agent. It helped us identify a performance issue in standard D365 F&O code and gave us a repeatable approach that may be useful on other customer projects.
D365 Trace Parser is a powerful tool for understanding what happens under the hood while a process runs. However, manually reviewing a trace can be slow and labor-intensive. The Trace Parser Agent performs much of that analysis and presents the findings in a report that is easier to review.
The Customer Issue
Users reported that entering Project Budget Revisions was extremely slow. We reproduced the issue in one of the customer’s internal environments, captured a trace, configured the Trace Parser Agent, and asked it to evaluate the trace for performance problems.
Within one to two minutes, the agent highlighted several areas for investigation, including customizations associated with Project Budget Revision objects. It then organized the results into an easy-to-read report for technical review.
Comparing the Customer Trace with a Baseline
To separate customer-specific behavior from standard application behavior, we captured the same process in an internal F&O environment. That environment contains demo data and no custom code in the affected area. We provided both traces, the customer trace and the Hummingbird trace, to the agent and asked it to compare them.
The comparison report highlighted performance issues in standard code and reduced the likelihood that the customizations were the primary cause.
We did not treat the AI output as the final answer.
We reviewed the X++ code identified by the agent and confirmed that the implementation details aligned with its findings.
Supporting a Microsoft Support Case
After validating the results, we used the report as source material for documentation supporting a Microsoft support case. Claude generated the initial draft, although Copilot could have been used for the same step.
Why This Matters for D365 F&O Consultants
This approach shortened the trace-analysis cycle, helped us distinguish customization concerns from standard application behavior, and produced evidence that could be reviewed before escalating the issue to Microsoft.
Ideally, Microsoft will provide a fix for the underlying issue. In the meantime, we have also used AI to explore potential customization-based options for improving performance. Any such change would still require normal technical review, testing, and validation before implementation.
THE TAKEAWAY
For end-users and consultants, the broader takeaway is straightforward: AI-assisted trace analysis can accelerate investigation, but the strongest results come from pairing the agent’s findings with a controlled baseline and expert review of the affected business process and X++ code.
Key takeaways
- Faster trace investigation: The Trace Parser Agent analyzed the customer trace within one to two minutes, identified areas for investigation, and organized its findings into a report for technical review.
- Baseline comparison: Comparing the customer trace with a trace from an internal environment containing demo data and no relevant custom code highlighted issues in standard code and reduced the likelihood that customizations were the primary cause. The team reviewed the identified X++ code rather than treating the agent’s output as conclusive.
- Validated support evidence: After confirming the findings, the team used the report to support a Microsoft support case. The approach shortened analysis, but any proposed customization changes still require technical review, testing, validation, and controlled comparison.