WHAT IS THIS FOR?

The Data Due Diligence Scorecard helps you to understand the quality of your data.

Make sure to fill the scorecard entirely to gain a complete overview on your data. This scorecard is a tool for practitioners that have already experience with AI and Machine Learning. Give us a shout if you struggle to ask some of the questions in the form. We are happy to share our views in a 10m call.

STEP-BY-STEP GUIDE

STEP 1 : GATHER DATA

Think about the data you have in your systems (E.g. CRM, log server, google sheet, etc.) and try to get a sample for each data source.

STEP 2: FILL IN THE SCORECARD

Go through each step and try to answer all the questions. Remember that this is an advanced scorecard, so don’t worry if you don’t have all the answers or you are struggling with a question for the moment.

STEP 3: UNDERSTAND WHERE YOU ARE

Count the number of topics for which you have less or equal to 2 answers. The quality of your data is indicated in the table below.

Score Data Quality Recommendation
7 → 8 ???? - poor create ASAP data collection processes
4 → 6 ???? - fair review your data processes
3 → 4 ???? - good start experimenting right now
1 → 3 ???? - great you are already done with data due diligence
0 → 1 ???? - excellent ready to scale up your models ????





      Gemmo's noise classification case study with Sonitus