Blog
Ideation
With a solid understanding of your users and data, it’s time to generate ideas and prioritize AI experiments. This phase combines creative ideation with designing experimentation to ensure that solutions are both impactful and feasible.
🎨Ideate: Creating Ideas and Selecting
Start by stating user needs, then challenge assumptions to create innovative solutions. While there are many variations of ideation frameworks (like Design Thinking, SCAMPER, Six Thinking Hats), they all follow these key principles:
Key Principles
Putting Principles into Practice: Mind Mapping
One effective way to apply these principles is through mind mapping, which helps organize both creative ideas (divergent thinking) and practical solutions (convergent thinking). Let's see how this works with a car buying example:
1. Start with Research Question
"How can we make buying a car feel like getting advice from a friend?"
2. Divergent: Explore Wild Ideas
Creative Ideas
Why This Could Work
Create a personalized car guide based on lifestyle 🚗
People trust recommendations that reflect their daily needs
Match cars like Spotify matches music preferences 🎵
Proven recommendation systems can be adapted to cars
Virtual test drive simulator with AI feedback
Combines real experiences with personalized guidance
There are no rows in this table
3. Converge: Pick Solutions
Focus Area
Solutions
Finding the Right Car
• Match cars to lifestyle
• Show similar cars others liked
• Learn from browsing patterns
Building Trust
• Clear recommendations
• Honest price comparisons
• Future cost predictions
Making it Personal
• Remember preferences
• Use plain language
• Adapt to feedback
There are no rows in this table
⚙️ ML Experimentation
Now that we have our solutions mapped out, how do we turn these ideas into working AI features? Let's bridge the gap between creative ideas and technical reality through ML experimentation.
ML Design Explained
Phase
Activities
Car Buying Example
Requirements Engineering
Define user/system needs, document technical constraints, and align specs.
Turn "friend advice" concept into technical requirements
ML Use-Case Prioritization
Rank ML applications based on impact and feasibility.
Start with browsing-based recommendations before adding style matching
Data Availability Check
Map data sources, assess quality, and identify gaps.
Confirm we have enough user browsing data
There are no rows in this table
🚀 Let's Connect
I'm always excited to discuss the intersection of AI and user experience, or explore potential collaborations.
✉️
🌎 Citizenship: US, Canada
📍 Location: Toronto, ON, Canada
💼
Want to print your doc?
This is not the way.
This is not the way.
Try clicking the ⋯ next to your doc name or using a keyboard shortcut (
CtrlP
) instead.