Photo by Anna Anisin.
Last week, I had the pleasure of attending the Data Science Salon conference in Seattle, Washington. It was my first time attending a Data Science Salon event and I was really impressed! Special thanks to the FormulatedBy team for the invite.
The event featured a great line-up of speakers and there was a diverse mix of attendees. The theme of the conference was “Applying AI and Machine Learning to retail and e-commerce.” Here are some of my takeaways from the 2019 Data Science Salon conference in Seattle:
- The accessibility of data science and machine learning today gives brick and mortar retailers the opportunity to compete online and online retailers to compete offline with local retailers. This accessibility is largely driven by Cloud computing’s opex model, the increasing amount of data being collected, and the availability of talent and knowledge.
- Data quality is still a significant challenge for organizations. Even in an industry where e-commerce and online markets drive sales, keeping data clean is still the large majority of the work.
- Like we’ve seen in most industries, the operation of machine learning systems isn’t getting much love. In many situations, data scientists have to stop working on solving business problems in order to check on the performance of a model or to refit it as it drifts. This is a challenge that is easily solvable through a well-designed MLOps platform and process.
- Architectures and algorithms may vary from use case to use case and sometimes from industry to industry. What remains constant is that you must have a talented and balanced team covering multiple roles that will enable you to answer your business question, build a solution to answer that question, and continue to answer that question correctly.
While retail hasn’t been a large industry focus for phData over the last few years, our model for machine learning is designed to support the entire machine learning lifecycle, regardless of industry. It was great to see companies working within the different lifecycle phases and reaffirming our understanding of where the challenges lie. I remain confident in phData’s Machine Learning team and their ability to help companies derive value from their data through Data Science and Machine Learning across all industries through their focus on the following core principles:
- Build models with production in mind
- Optimize data scientist productivity
- Move machine learning from powerpoint to production
- Keep models performing at their prime
After a great experience in Seattle, we’re excited to announce that we’ll be sponsoring the Data Science Salon event in Austin, Texas on February 18th-19th, 2020. Additionally, I’ll be giving a talk at the event so be sure to block off time for MLOps: The Assembly Line of Machine Learning. If you want to talk more about the upcoming event, let’s chat! We hope to see you there!
Chief Machine Learning Architect
Connect with me on LinkedIn