Agile AI Assistant

Agile AI Assistant for Jira

Inspiration

We recognized a gap in the market for tools that ensured consistent adherence to Agile principles across an entire organization. Many teams struggled with ticket clarity and the adoption of best practices. We envisioned a tool that could seamlessly weave AI and Agile together, enhancing productivity and precision within Jira Cloud.

What it does

The Agile AI Assistant for Jira refines user stories, tasks, and bugs. With a single click, users can align tickets with Agile principles. The application tailors content to meet Scrum standards, converts acceptance criteria into actionable checklists, and aids in the seamless adoption of Agile methodologies company-wide.

How we built it

The application has been created in the Forge environment. We utilized the cutting-edge OpenAI model and integrated it with Jira's platform. Communication between the Forge application and OpenAI is done using REST API calls. We are collaborating closely with Agile experts to adapt content (and verify it) to fit Scrum and other Agile standards, ensuring a tool that's intuitive and efficient.

Challenges we ran into

At this stage of development, getting to know the forge environment and overcoming technical difficulties is challenging. The main problems we have encountered are timeouts during processing and crashes during generation. However, we are happy that we were able to overcome these difficulties. Selecting the proper service to give us the expected results. We have tried different services, and we found one that, for today, gives us the best results. Another aspect was finding a balance between AI automation and user customization. We also faced hurdles in ensuring our AI truly understood and adapted to the nuances of different Agile methodologies and practices across diverse industries.

Accomplishments that we're proud of

Successfully creating an AI that refines ticket content and educates users on best practices. Achieving seamless integration with Jira and witnessing its adoption across varied organizations, enhancing their Agile workflows.

What we learned

The amount of knowledge we have had to digest recently is remarkable. Forge, AI services available on the market and utilizing them for our project. The depth of nuances in Agile methodologies is vast. We learned the importance of iterative feedback and continuous improvement, much in line with Agile principles, to perfect our tool. Collaboration with end-users provided invaluable insights. Focus on deliveries, find workarounds, and ask the team for help if you have encountered blockers. Our motto is "United we stand, divided we fall".

What's next for the Agile AI Assistant?

Expanding our AI to cover more Agile frameworks, enhancing the user interface based on feedback, and possibly integrating with other AI services to compare results. The next step is pretty obvious. We want to start training models. Not only using external services (we do not plan to drop them) but tuning answers to be more accurate. The world of AI is changing so much and has never been faster than it is now. New models are released daily, so keeping your knowledge up-to-date is a necessity.