Logo

Register Now

View My GitHub Profile

Machine Learning in Production

Do you feel like you or your company are not shipping Data Science projects fast enough? Do your ML projects get stuck because there aren’t available developers to bring them to production? Do DevOps and SWE conversations sound like Klingon to you?

Then this course may be for you!

Join us to:

Our training offers:

This course has provided me with a bunch of tools (kind of a Swiss Army knife) that have made my work day easier. Now I have the basics needed to build a proper scaffold on which to develop my ML projects, so that they can be easily placed in production. It’s a must!

– Marta Dies, Data Scientist, ML in Prod 2019 alumnus

The course was really useful to understand what tools are needed in order to put models into production

– Cristian Pachón, Data Scientist, ML in Prod 2019 alumnus

Register Now - February 2020

Most Machine Learning projects never see the light of day

With the advent of internet-scale data gathering, powerful big data platforms and new computing paradigms, most companies have embraced the Big Data and AI revolutions. As a result, billions have been invested to build all-mighty, data-powered features and services, in the hope of getting a competitive edge ahead of slower competitors.

But alas, now that the dust is settling, a growing number of companies are starting to ponder: is this huge investment paying off? Is my bright, PhD-holding, Data Science team delivering on its promises?

The sad answer is that in many cases, our Return on Investment is not great. And it is not necessarily our fault.

Data Science is hard to bring to production

If you have been in the Data Science business for long enough, the following situation will likely sound familiar:

At this point you realize that you are not quite sure about the answer… How will you deploy the model? What infrastructure will it run on? The prototype kind of works, but will it generalize well to completely unseen data? How will you re-train it with a larger dataset? How will you work with other peers to iterate on the current model?

But you say something like “a couple of Sprints?” and hope for the best… only to realize 6 Sprints later that the model isn’t yet life and everyone is starting to wonder whether the time investment is worth it.

Get better at shipping Machine Learning models

Luckily at this point there are a number of people and companies (including us) who have been facing these problems for a few years. In our experience, the key elements for successful ML deployment are:

  1. Working with iteration and deployment in mind, from the start
  2. Using tools and practices from Software Development
  3. Knowing the basics of Data Engineering and DevOps to become more autonomous

The course will consist of theory and practical hands-on sessions lead by our four instructors, with over 20 years of cumulative experience building and deploying Machine Learning models to demanding production environments at top-tier internet companies like edreams, letgo or La Vanguardia.

Our training

Syllabus and schedule [Tentative]

Warning: We are trying to come up with the most relevant content so this syllabus is still work-in-progress, which means it is susceptible to change.

Instructors

Bernat Garcia Larrosa: Degree in Mathematics and Industrial Engineering. Former Data Scientist at Diari ARA and LaVanguardia.com. Current Director of Big Data at yaencontre.

Aleix Ruiz de Villa: PhD in Mathematics. Founder of the Barcelona Data Science and Machine Learning Meetup, cofounder of the Barcelona R Users Group and cofounder of BaDaSS. Former Head of Data Science at LaVanguardia.com, SCRM (Lidl) and Onna. Currently Chief Data Science Officer at Flaps.io and machine learning/causal inference consultant.

Tristana Sondon: PhD in Computational Physics. A former Academic Researcher, she has been working for the past 5 years in Management consulting on areas of Data Science for banking, retail and ecommerce companies. Currently a Data Science Lead in the area of Revenue Management and Dynamic Pricing.

Arnau Tibau Puig: PhD in Electrical Engineering and flamenco lover. Former Principal Engineer at @WalmartLabs, Lead Data Scientist at Quantifind, both in the San Francisco Bay Area, CA. Current Head of Data Science at letgo, Barcelona.

Capstone and more info!

If you want to have a look at the capstone we are going to work with, check out this github repo.

If you are curios and want to know more about what we are going to talk about, have a look at this Medium Post.

Registration information

Dates:

Venue: Llibreria Laie, C/ Pau Claris, 85, 08010 Barcelona

Registration Fees:

Language: English

Contact: bcn.mlinproduction@gmail.com

Register Now - February 2020

Requirements (Important!)

Applicants should have some experience with:

Students must bring a laptop equipped with:

Important: Please follow the environment setup instructions prior to the beginning of the course to ensure a smooth learning experience.

Past editions

We organized a June 2019 edition of this training in Barcelona, Spain. 100% of our post-course survey respondents said they would recommend it to a friend, and we have incorporated their suggestions on how to make 2020 edition even better!

alt text

alt text