I'm a software developer at a software company, and our organization is encouraging us to find practical ways to use AI in our work. We use tools such as Claude Code, Jira, GitHub, and Datadog, and I'm looking for project ideas that could improve development workflows, team processes, or day-to-day productivity. I'd especially appreciate examples that are genuinely useful rather than AI for its own sake, since I'd like to take on a visible project and contribute more to the team.
3 Answers
Start by identifying a repetitive process that already causes frustration, then see whether AI can reduce the manual work. For example, you could build a tool that summarizes Jira tickets and GitHub changes into a release note draft, identifies tickets that are missing acceptance criteria, or turns bug reports into a consistent template. A small project with measurable time savings will usually get more attention than a large, vague AI initiative.
Incident and support workflows are another good area. You could use deployment data, Jira issues, GitHub history, and monitoring alerts to create an incident summary, suggest related past incidents, or draft a post-incident report for an engineer to review. Keep a human approval step in place, and measure things like reduced triage time, fewer repeated investigations, or faster handoffs.
Before choosing a solution, talk to a few teammates and ask what they do repeatedly every week that they dislike. Possible ideas include generating test cases from requirements, reviewing pull requests for project-specific conventions, searching internal documentation, or detecting inconsistencies between code and documentation. Avoid presenting it as 'we need to use AI'; frame it as solving a known problem and compare the AI-assisted process with the current one.
That makes sense. I was initially focused on finding an impressive idea, but measuring a real improvement in an annoying process would probably make a stronger case.