Product Development Framework Documents

Free documents for re-use to develop (ML/ AI) products

As part of my work at Taimaka, I developed several technical documents that can be leveraged by other organizations and companies to make their processes and product development easier and more structured. 

These documents include:

1. User Research Templates - frameworks + tips on conducting helpful user/ customer interviews

• Recommended use: Use this first if you still don't have a great understanding of what your core user/ customer problems are or deeper segmentations. 
• Example questions you might have which mean you should conduct some user research: How exactly are people using my product? What kinds of people find my product useful or not useful? What parts of the product are working for people or not? Are there more specific people we should be targeting to improve our product? 

2. Product Prioritization Sheets -  templates to quantify product prioritization decisions and track progress 

• Recommended use: Use this sheet if you have a bunch of ideas generated but aren't sure which ones to tackle first, or have lots of projects but struggle to track progress consistently
• Example questions you might have: Which project should we focus our time/ energy on first? What projects can be deprioritized for now? What's the status of project X and who last worked on it? 

3. Product Spec - formatted document + guidance for technical product development 

• Recommended use: Fleshing out detailed product changes for a new feature/ program
• Example questions: What exactly do the engineers need to build to make the product viable? What are the exact tenants of the program we need to ensure are consistent across all platforms/ sites?

4. Model card template - for development and deployment of safer AI/ ML models

• Recommended use: When developing AI/ML models, understanding the risks of implementing them + clear documentation so there is understanding across the team what model does what and how they are used
• Example questions: What are the risks associated with building the model we want to implement? What do we need to consider when picking a model or data sources?