1.  2017, founded
2. Data centric deep learning
3. 
   
4.  how businesses without giant data sets to draw on can still join in the AI revolution.
5. LandingLens to help manufacturers improve visual inspection with computer vision

So the challenge is, can you get an AI to work with a hundred images?  

Challenge of smaller data set

~ hundred examples

 noisy data

= the wrong way 

 be customer-led or mission-led, almost never technology-led.

1. digital transformation so there’s some data. 
2. Make AI work on hundred images?  
3. data is never clean and perfect. It’s a journey. 
4. prioritize what data to collect.
5.  smaller data sets == cleanly labeled and carefully curated.
6.  noisy data has a huge impact on performance.
7. label the data consistently = very important
8.  Let’s not mess with the code anymore.
9. improve the quality of the data
10.