AI-Testground

This repository contains a list of questions we want to ask AIs to see how they can support us.

You are eager to read the results?

Software Development Process overview

This process overview lists all the things that happen during the development process. So we can have a look at it and feel inspired.

Task arrival

What is done in this state?

• The task arrives, probably as an email.
• The task is added to the Kanban board using a PowerShell script.
  • The duration of the task is estimated based on the subject.
  • The email is extracted, added as an attachment, and the last main section of the email is included as a description.
  • The task is added to a suitable column based on rules.
  • Classification of the task as support/bug/feature or internal development.

Improvement ideas for AI:

• Rewriting incoming emails
  • The subject of incoming emails is often not in line with the content, making it difficult to get an overview in the Kanban board. Create a new, appropriate subject?
  • Due to the inaccuracy of the subject, our task duration prediction is negatively impacted (can AI contribute here? e.g., by properly formulating the subject)
• Improving tasks
  • If AI can choose a suitable subject for tasks, the Kanban board overview will improve.
  • Check in the accumulating tasks if we already have a similar task from someone else?
  • Automatically create subtasks for the tasks.
  • Generate a standardized task description (Where is the problem? What exactly is the problem? Which customer does the system affect?)
  • Check if all important information is included in the content.

Handling of support tasks

Support tasks can have a very high variance, but there are also the usual suspects for which we already have a high degree of automation.

The challenging tasks are very knowledge-intensive and must be solved cleanly. AI cannot help here.

Handling Development tasks

1. Requirement collection

• Can we use AI to find out who is affected by a task?
• Automatically suggest solutions?

2. Planning and Design

• Propose an architecture?
• Write user story, use cases or functional specifications?
• Design the user interface?

3. Implementation

• Help us find out if someone is following the coding standards.
• Which development tasks can be done within the scope of our tech stack?
  • C#, MS SQL-Server, PHP, MySQL, TypeScript, Javascript, Powershell, Python, Rust
• Usability for performance optimization, e.g., in SQL
  • Rewriting a scalar function into a view

4. Testing and Quality Assurance

• Develop a test plan?
• Write unit tests?
• Identify and fix bugs?

5. Deployment and Release

• Can AI create scripts to help us automate deployment and release?
  • GitHub Actions
  • Powershell
  • Docker

6. Maintenance and Support

• Writing user documentation
• Writing training materials
• Suggest solutions for identified problems. (Proposal for immediate measures)
• Proofreading of emails and documentation materials (removing grammar and spelling mistakes)