Tracewell
Bug reports that arrive ready to fix.
An automation concept for software teams: AI asks customers for the missing details, gathers logs and recent activity, finds duplicates, estimates severity, and files a complete ticket in Jira or Linear for the team that owns the problem.
- Services
- AI automations
- System integrations
- Custom software
- Technology
- Python
- FastAPI
- Next.js
- TypeScript
- PostgreSQL
Steps to reproduce, the device, the user’s last 10 actions and the matching stack trace
1 / 7The challenge
Customers report “it doesn’t work” with no details. Support forwards the message, and a developer spends an hour finding out which page, which browser and which error before any fixing starts. Meanwhile, six other customers report the same bug in different words.
The approach
Tracewell reads each report and asks the customer only for what is missing: the page, what happened, the browser and a screenshot. It pulls the matching logs, the error with its stack trace and the user’s last actions, then groups the report with others caused by the same error to show how many customers are affected. It drafts a full ticket for the team that owns that part of the product, and a lead confirms the priority before it is created.
Designed around
- Follow-up questions to the customer, in seconds
- Matching logs, errors and recent user activity
- Duplicate detection with the number of customers affected
- A severity estimate with the reasons behind it
- Complete tickets in Jira or Linear
- Assignment to the team that owns that part of the product
- A lead confirms the priority before the ticket is created
More AI & automation projects
AI & automationAnime Creator
Consistent characters, from the first page to the last.
AI & automationKestrel
Find the businesses that need you.
AI & automationWalkaround
No more “that scratch was already there”.
AI & automationHala
Answers in seconds, even at 2am.
AI & automationShiplog
Release notes, written from the pull requests.