The conference was held at the breathtaking Maison Internationale in Paris. Louis Dorard, the founder of conference, set the tone for the day with his talk on Dymistifying Machine Learning. He explained what it is, how it works and where predictive APIs come in.

Maison Internationale in Paris

David Jones presenting at PAPIs Connect 2015. Photo credit: @VeroColbert.

I had the pleasure of being invited to present too. My talk was titled "Machine Learning & Ecommerce". Here are some key take aways from my talk:

Machine Learning & Ecommerce Pair Well

We got a 71% increase in revenue with a wine website by using even the most basic of machine learning features. Companies like Amazon have been using ML in ecommerce for years, this is because it delivers good results.

Small Data Can Be Enough

You may not need as much data as you think to use machine learning effectively. The key is quality. With ecommerce specifically you're looking for anything that links customers and products.

Predictive APIs Remove Barriers to Entry for Machine Learning

Machine learning comic presented about barriers to entry.

It wasn't too long ago that to use machine learning in practice you needed a team of experts. The skills required are quite deep. But Predictive APIs make it feel like you have your own team of ML experts and ultimately lower the barriers to entry with building predictive apps.

Business Leaders Should Start ML Projects

Most people don't use machine learning right now yet the benefits can be tremendous. Business leaders need to have machine learning on their radar or be starting a ML project soon. Start with the basics but start now and get your foot in the door early, the timing is perfect because the benefits are currently under capitalised.

Here are the slides from my talk.