Finding use cases in the company

1. Analysis: Where’s the problem?

The best place to start when looking for use cases is with your daily pain points. Ask yourself and your team: Which tasks are particularly time-consuming, repetitive, or unnecessarily complex?

This is exactly where the greatest potential for AI support lies, whether it’s creating reports, drafting emails, or processing data.

2. Think in categories & gather ideas

To structure your brainstorming, it helps to think in terms of AI’s core strengths. Collect potential use cases in these five categories:

• Creating text: (e.g., social media posts, blog articles, job postings)

• Summarizing content: (e.g., meeting minutes, long emails, studies)

• Analyzing data: (e.g., evaluating customer feedback, preparing spreadsheets)

• Developing ideas: (e.g., brainstorming for campaigns, naming)

• Researching information: (e.g., market analysis, topic research)

3. Conduct departmental workshops

The best ideas come directly from the employees who experience the processes on a daily basis. Therefore, organize short, focused workshops (approx. 60–90 minutes) with various departments such as Marketing, HR, or Sales. The goal is to jointly identify the biggest time wasters and areas of potential within each department.

Which tasks occur daily, weekly, or monthly? Which tasks take up time?

4. Prioritize quick wins (low-hanging fruit)

Not every idea should be implemented immediately. Evaluate the collected use cases based on two simple criteria: What is the potential benefit (e.g., time savings, quality improvement)? And how easy is it to implement using InnoGPT?

Start specifically with the "quick wins"—that is, the cases that promise high benefits with minimal effort.

5. Define pilot use cases per department

Select a top candidate from each workshop as a pilot use case. Define this use case clearly and in a way that everyone can understand: What is the exact goal? What does the specific workflow with InnoGPT look like? And what is the optimal prompt to achieve the desired result?

Build the assistant together; this will also provide a direct learning experience for the Prompting of assistants.

6. Share successful use cases as assistants

If a use case proves valuable, it shouldn’t get lost in a drawer. Use Assistant function in InnoGPT to build a finished assistant from the successful prompt and workflow. You can then share this with the entire workspace so that every colleague can use the proven use case with just one click.