You don't need AI; You need to know what your data is

I've had a rant brewing in me for 5 years, which I finally articulated into a semi-coherent talk. The nice folks at PyConZA allowed me the opportunity to talk about what's been on my mind.

93% of data science projects never make it into production. Many of us in the field of AI and Data Science feel the impact of this — the sheer futility of all our efforts.

First I talked about how we ended up in this mess. Spoiler Alert: We were misled. Next, I provided my opinion on how to get OUT of it. It's time to put down those neural networks, and pick up some data governance tools. It's time to stop hiring data scientists and instead hire data engineers.

Join me on a journey to go back and rework some of the foundations so that our Data Science efforts can begin to make an impact. This talk is complete with Wardley maps, sparkle GIFs, and Ru Paul memes.

Audience: Anyone who works in data science or runs a data science team.

What will they get out of it:

  • Increased Skepticism on what the tech zeitgeist proposes.
  • A list of many of the practices we follow in Data Science which are ill advised.
  • What we can change to ensure we actually make an impact.

About the author

Jade Abbott profile picture

Jade Abbott

Jade Abbott

I’m the ML lead at Retro Rabbit where I’ve worked at every end of putting an ML model into production. By night, I lead Masakhane, a grassroots open research movement for African language technologies Read more from Jade Abbott...