A convincing project shows more than an impressive output. It explains the problem, your approach and how you checked the result.
Choose a real, manageable task
Start with something familiar: organising public research, preparing a meeting brief, drafting a sourcing plan or comparing information. Use public, synthetic or properly authorised data.
Write the brief before the prompt
Describe the intended audience, input, desired output and constraints. Define what a useful result looks like. This gives you a basis for judging the tool’s response.
Document what you changed
Save the initial approach, revisions and the reasons behind them. Did you add context? Narrow the task? Specify a format? The decisions tell a stronger story than a list of tools.
Check the output
Review accuracy, relevance and completeness. Verify factual claims against suitable sources. Record mistakes and explain where human review remains necessary.
Tell the story in five parts
- Problem: What needed to improve?
- Approach: Why did you choose this workflow?
- Execution: What did you build or produce?
- Review: How did you check quality?
- Learning: What would you improve next?
Describe measured results only when you have actually measured them. Clear thinking and honest limitations make a project easier to trust.