search

zikrya / TaskMaster-AI

https://www.devliftoff.com/
0 stars 0 forks source link

[OPERATION] Research Necessary Scrum and Agile methods to fine tune AI #36

Open zikrya opened 5 months ago

zikrya commented 5 months ago

@guapg

Research how PM using agile and scrum create tickets, as in the thought process, what they use to think of tickets, how they organize them, etc, and list it down, so I can use that info to train the AI to recreate that the best it can

guapg commented 5 months ago

Agile methods have an emphasis on iterative development, continuous collaboration, and adaptability which is key to helping fine-tune AI. The goal is for the AI to be able to automate feedback loops and streamline communication among teams. Scrum, with its structured framework, includes defined roles, artifacts, and regular events (sprints and retrospectives). While thinking about the way a PM goes about their thought process, the end goal is to provide constant feedback.

  1. Summarizing the End Goal Overall project goals, roadmap, and how each ticket will help for the bigger picture. Reviewing product backlog

  2. Define User Stories Writing requirements into user stories, helping keep the focus on user value. "As a [type of user], I want [benefit] so that [some reason]." The end goal is that the team builds features actually needed by its users

  3. Simplifying User Stories into Manageable Tasks Break each story into smaller tasks during planning sessions

  4. Prioritizing and Organizing Tickets Urgency, dependencies, value towards the business, and how they align with the goals. MoSCoW (Must have, Should have, Could have, Won’t have) for prioritization Tickets are organized into sprints and then checked for review

  5. Calculating Effort The development team provides input to estimate the effort necessary for each ticket with methods like Planning Poker / T-shirt Sizing

  6. Tracking and Adapting Scrum / Kanban boards used for ‘to do’, ‘in progress’, and ‘done’ Daily stand-ups help the team stay on track and adapt/adjust to necessary changes. PM updates everything accordingly

  7. Reflection After each sprint, PMs lead a sprint review and retrospective to iterate what went well and what could be done to improve. This helps future ticket handling. Tailoring our AI to perfect these factors will help ensure continuous improvement.

zikrya commented 5 months ago

@guapg

Can you formulate on ways we can use these practices in our own AI?

What practices best highlight how our tool can we be used? How could we integrate stories into our AI? How do you usually break stories in agile/scrum?

Best, Zikrya

guapg commented 5 months ago

Our AI can assist users by creating user stories and suggesting templates and autofill categories based on similar projects they have performed in the past. Users input requirements and then our AI can format them into user stories

Breaking each story into more manageable tasks and helping with time estimation based on previous projects. Data can be collected through a timer on other projects to help with this.

AI can prioritize tasks with MoSCoW framework from input from stakeholders and project goals

AI can organize tasks into sprints, helping balance workload and dependencies → targeting the best sprint length and the structure alone with it

Planning Poker can also help the AI estimate the efforts required

Our AI can update Scrum/Kanban boards automatically and notify team members of status updates

AI can also help summarize daily stand-ups

Help with Retrospectives as it can continuously analyze performance in each sprint and the data along with it to help future sprints.

Breaking the stories consists of understanding the story, identifying subtasks, collaboration with the team, prioritizing tasks, and estimating effort