The document discusses the importance of understanding data structures when designing products. It notes that product designers and data scientists both aim to reduce friction. Their work intersects as user experience depends on the underlying data architecture. Different data structures like relational databases, graphs, and knowledge graphs are suited to different problems. Case studies show how graphs power applications like image recognition and last-mile delivery by connecting product, inventory, logistics and other data. The document proposes a data thinking prototyping framework to map business problems, data models, value opportunities and applications when considering new solutions.
Unraveling Multimodality with Large Language Models.pdf
Neo4j Innovation Lab – Bringing the Best of Data Science and Design Thinking Together
1. Design Thinking 💙 Data-Science
Connections, August 25, 2021
The importance of understanding data-
structures when designing products
Alessandro Svensson, Head of Neo4j Innovation Lab
3. Neo4j Innovation Lab
The intersection of product design & data-science
01
02
03
04
05
Introduction to the Neo4j Innovation Lab
Product designers and data-scientists
Data structures & the data value chain
Real world examples
Data Thinking Prototyping Framework
Today’s Agenda
5. Help companies accelerate innovation and digital transformation
through graph thinking.
What we do
We generate and prototype graph projects together with
customers and prospects.
How we do it
Expected Outcome
To provide a deep understanding of graph thinking and the
new possibilities in innovation and digital transformation that is
enabled by adopting Neo4j and graphs.
Innovation Lab Sprint — 3 day workshop
Methodology
Neo4j Innovation Lab
8. Storyboarding/
Mockups
Identify Stakeholders / Synopsis
Functional – Emotional – Social Needs
Sprint scope/
Define Goals
Define Target Use
Case
Reference Graph
Data Model
Executive Presentation
Solutions Roadmap
Graph Related
Questions
Source data to
populate model
Build Query /
Pipeline
Solutions
Framework
Build Prototype/Wireframes
Behind the Scenes
On Stage
Methodology
9. Neo4j Innovation Lab
1. Defining Use Case
UX/UI Lead
Data Scientist
Head of Innovation Developer
CIO
Business Analyst
Head of Customer Success
10. Neo4j Innovation Lab
2. Data modeling
Adam Cowley, Neo4j
Field Engineer
Director of AI
Data Scientist
Eric Monk, Neo4j PS-
consultant
Innovation Lab
Leader
18. Neo4j Innovation Lab
• Data Scientist
• Data Analyst
• Data Engineer
• Data Architect
• Data Storyteller
• ML Scientist
• ML Engineer
• BI-Developer
• Database Administrator
• Technology Specialized Roles
• Etc…
Data Workers
https://towardsdatascience.com/10-different-data-science-job-titles-and-what-they-mean-d385fc3c58ae
19. Neo4j Innovation Lab
Multi-faceted area of work
Data
Capture
Data
Cleaning
Data
Modeling
Data Analysis
Visualization
Deployment
& Scale
Evolve
20. Neo4j Innovation Lab
Compute complexity
Reduce Friction
On stage
User Experience
Underlying Data
Behind the scenes
Where design meets data work
21. Neo4j Innovation Lab
Compute complexity
On stage
User Experience
Underlying Data
Behind the scenes
https://dribbble.com/shots/15633889-Investment-Mobile-Apps
https://dribbble.com/shots/14951197-Social-app-design-Light
Investment App Social Media App
22. Neo4j Innovation Lab
On stage
User Experience
Underlying Data
Behind the scenes
https://dribbble.com/shots/15633889-Investment-Mobile-Apps
https://dribbble.com/shots/14951197-Social-app-design-Light
Investment App Social Media App
25. Neo4j Innovation Lab
Data Value Chain
Raw data Captured data Organized data Processed data
💡
Conceptual, no value
26. Neo4j Innovation Lab
Data Value Chain
Raw data Captured data Organized data Processed data
💡
Conceptual, no value
🧐
Potential value
27. Neo4j Innovation Lab
Data Value Chain
Raw data Captured data Organized data Processed data
💡
Conceptual, no value
🧐
Potential value
🤑
Valuable
28. Neo4j Innovation Lab
Data Value Chain
Raw data Captured data Organized data Processed data
💡
Conceptual, no value
🥳
Very valuable
🧐
Potential value
🤑
Valuable
29. Neo4j Innovation Lab
Data Value Chain
Raw data Captured data Organized data Processed data
Power analytics
Power applications
💡
Conceptual, no value
🥳
Very valuable
🧐
Potential value
🤑
Valuable
30. Neo4j Innovation Lab
“A database is an organized collection
of structured information, or data,
typically stored electronically in a
computer system.”
Tools for organizing data (databases)
— Wikipedia
36. Neo4j Innovation Lab
Databases
Potential Raw
Data to Capture
Data Warehouse
Data Lakes
This makes sense for
connected problems i.e
recommendation engines
💡
Graph
Architecture:
Choosing the
right tool for
the job
51. Neo4j Innovation Lab
On stage
User Experience
Underlying Data
Behind the scenes
https://dribbble.com/shots/15633889-Investment-Mobile-Apps
https://dribbble.com/shots/14951197-Social-app-design-Light
Investment App Social Media App
53. Neo4j Innovation Lab
Data Thinking Prototyping Framework
Business Problem/Opportunity
What is the business problem/opportunity you want to
solve for with graphs and connected data, and why is it a
relevant problem to solve?
Data Model &
Ingestion
What does your data-model look like,
and how should you ingest the data?
Underlying Solutions
Architecture
How will graphs be fitted into your overall
architecture?
People, Teams &
Skills
What are the stakeholders, organizations
and skillsets required to be successful?
“Money” Queries
& Algorithms
What are the queries and algorithms
relevant to extract value
Value unlock
What is the value that will be unlocked
due to connecting your data?
Applications &
Services
What are the tangible touch
points?
BUSINESS
ORGANISATION
Data Capture &
Strategy
What are the essential data-points and behaviors
relevant for your use case to capture?
DATA
54. Neo4j Innovation Lab
3 Key takeaways
Your apps are as good as the data that powers them.
What happens “behind the scenes, shows up on stage”.
👆
Use data-thinking to trigger your imagination. Mapping data
strategy gives you a lot of ideas for function.
✌
You don’t have to know every technology in detail,
a basic understanding of data-structures goes a
long way in choosing the right tool for the right job.
🤟