Automating Standup and Ticket Updates with a Custom AI Tool
The author built a custom AI-powered tool called DevTrack to automate daily standup updates and ticket management, saving over 20 minutes per day and automating 94% of work item updates.
Why it matters
This project demonstrates how custom AI-powered tools can automate repetitive tasks and improve developer productivity, reducing manual work and freeing up time for more strategic activities.
Key Points
- 1Developed a Go daemon that watches Git repositories and uses NLP and LLMs to automatically update tickets and generate standup notes
- 2Personalized the AI to generate updates in the author's own communication style
- 3Reduced manual work from 11pm Jira sessions to nearly zero, with 4 out of 5 standups pre-written
- 4Initial setup of the Ollama LLM model can be challenging, but the author is working on improving the onboarding experience
Details
The author initially used a Google Sheets hack to automate standup updates, but it still required manual work. To fully automate the process, they built DevTrack, a custom tool that integrates with Git, NLP, and LLMs to automatically update tickets and generate standup notes. DevTrack runs locally, using the Ollama LLM for the AI processing. It extracts context from commit messages, associates them with open work items, and posts updates to the project management tool. The author has been using DevTrack for 6 months and has seen significant time savings, with 94% of work items updated automatically and 4 out of 5 standups pre-written. The personalization feature, which tailors the updates to the author's communication style, has also been well-received by their manager.
No comments yet
Be the first to comment