Mike Skarlinski, Manager, Data Science for WW (The New Weight Watchers), will discuss the challenges (and potential solutions) that data science teams face in operationalizing models, building shared infrastructure, and on-boarding new members.
This is the major underlying theme for his presentation Going Beyond Big Data – Taking Machine Learning (ML) to the Next Level – at the upcoming DATAx Conference, taking place on November 6-7 in New York.
After reviewing some of the unique challenges at his own organization, he will talk about how designing and building a machine learning deployment framework that can facilitate the rapid growth of a team, and increase project velocities.
In Skarlinski’s session on November 7, attendees can expect the following three major takeaways:
- Why it’s important to take stock of an organization’s technical needs before developing data products; many times simplicity can be chosen over the need for scale.
- Learn that effective data products often go “beyond the model”, and empower data scientists to develop the end-to-end solutions can result in faster and more flexible development processes.
- How to use and develop a framework for machine learning that can facilitate data scientist on-boarding and greatly reduce the incremental effort of new model deployments.
Skarlinski recently identified that it can be challenging to integrate data products into existing infrastructure and practices. Organizations often want to “sprinkle” machine learning into their applications as an additive measure, without realizing that effective ML systems need to be deeply integrated into the technical stack and product leadership of an organization. Solutions often require extensive internal education, senior leadership buy-in, and an experienced technical team.
Another finding Skarlinski made was that a consultative approach can facilitate building new data products within existing systems. In this way, necessary architecture and product changes are proposed to teams simultaneously, and at a high level. This makes sure leadership, product, and technical folks are all aligned before diving into new data product developments.
In his opinion, the industry is really lucky to have quality open-source research contributions coming from major tech brands like Microsoft, Google, and Facebook. Their work enables smaller teams to see state of the art approaches and experiment with implementations.
Skarlinski offered up some great advice for recruiting and maintaining top talent and for leaders looking to build high-performance teams.
“Retaining and recruiting top talent is of paramount importance. Team members need to be kept interested through new challenges and increased responsibility. In addition to making the best hires, I think that it’s equally vital to build a team culture which allows for junior members to grow and learn effectively from top talent.”
When asked about his current Artificial Intelligence (AI) or Machine Learning (ML) initiatives, Skarlinski shared that his data science team has been busy developing Primrose, their open-source machine learning framework, for building and deploying models. They have utilized their framework to build data products across the organization including real-time membership models, recipe recommenders and personalized social-media feeds for WW’s social network, Connect.
To hear more from Michael Skarlinski, reserve your place at the upcoming DATAx New York Conference on November 6 -7. Act fast to secure your ticket at the discounted early-bird price. [LEARN MORE AND REGISTER NOW]
Visit Argyle Executive Forum's 2019 CIO Think Tank: How to Optimize Multiple Cloud Platforms in New York, NY on Sep 26, 2019