We are excited to announce that we’ll join forces with Venture Beat for their Transform 2021, one of the most important events of the year for enterprise technical leaders on how to implement applied AI.
The Data Science Salon will be hosting the data science in the enterprise track at the conference, covering state-of-the art AI and machine learning applications in some of the hottest industries, including finance, retail, healthcare, media, advertising & entertainment.
Chief Technology Officer at FreeWheel
Senior Director of the Search Sciences Team at Nike
Senior ML Engineering Manager, Personalization Lead at Peloton
Sr. Director, Head of Analytics at Roku Inc
Head of ESG Product at Dun & Bradstreet
Head of Data Science & Engineering at Condé Nast
Senior Director Enterprise Data Strategy and Engineering at CVS Health
CEO & Founder at Mobius Labs
Field CTO - Telecom Strategy at Data Stax
The data science salon is a unique vertical focused conference which grew into a diverse community of senior data science, machine learning and other technical specialists. We gather face-to-face and virtually to educate each other, illuminate best practices and innovate new solutions in a casual atmosphere.
11.30AM – 11.40AM
Introduction & Housekeeping
11.40am - 12.10pm
Building A Smarter Media Marketplace: Machine Learning & AI Demystified for TV Advertising
Yuling Ma – Senior Vice President of Engineering at FreeWheel
With data as the connective tissue of the media advertising business, a new way of thinking is needed to solve the unique challenges of applying data, machine learning and artificial intelligence across the TV advertising ecosystem. In this special presentation, Yuling Ma, Chief Technology Officer, FreeWheel, will walk through the latest thinking from FreeWheel.
12.15pm - 12.45pm
Considerations for Successful Model Management
Saira Kazmi – Senior Director Enterprise Data Strategy and Engineering at CVS Health
AI/ML is top of mind for leaders across academia and the industry. Many have tried and failed several times before successfully deploying models that bring real business value or have an impact. Even after a model is deployed, constant maintenance and monitoring is required to ensure that the model is relevant and makes good decisions that are still applicable to the changing business environment. Key factors leading to successful deployments include: * Availability and maturity of data to build a model and availability of this data when making real-time decisions * Thorough understanding of the business problem * Understanding of nuances leading to data variability * We can often overestimate what models can do – a simple test “Is the Candidate task simple for a human to solve?” can help. * Automation will lead to ROI (the problem is large enough to automate) * Mechanisms are in place to track the algorithm performance * There is a way to provide feedback on the solution and model decisions * The model is kept alive (refreshed with new data at regular intervals
12.50pm - 1.20pm
Innovate with AI & ML: Achieving the Data Driven Enterprise with DataStax & Cassandra
Reed Peterson – Field CTO – Telecom Strategy at Data Stax
AI systems exhibit learning, planning, reasoning, decision making, and problem-solving. AI is a step deeper than Machine Learning and, when supported by a modern data stack, brings tremendous value to both enterprises and their customers. Join this session to: * Learn the ideal architecture & features required to deliver AI/ML solutions – including real time data, scalable infrastructure, intelligent replication, relevant data & dynamic provisioning *Discuss how the process should work along with some of the key challenges & pitfalls that limit success *Walk through ML & AI use cases and the key ways take advantage of them in your business
1.25pm - 1.55pm
Scaling and Transforming Stitch Fix’s Visibility into What People Will Love
June Andrews – Data Science Manager of Style Discovery at Stitch Fix
A central component of Stitch Fix’s ability to match clients with clothing they will love–is the data-driven curation of our expansive inventory. Knowing ahead of time how well our inventory will perform not only reduces costs associated with missing the mark on what clients will love; it also provides great insight into the inventory that will be most successful with a growing client base in expanding sales channels. Here we present the history, lessons learned, adjustments in the face of 2020’s historic challenges, and important milestones in developing a recommender system focused on inventory curation. This recommender system, Style Explorer, predicts what items our clients will love–often before those items have even been fabricated. Providing Style Explorer as a tool available at all stages of the Stitch Fix vertical supply chain has de-risked and augmented processes ranging from design and fabrication to purchasing. In the process, it has transformed and scaled our visibility into what people will love.
2.00pm - 2.30pm
Improving ML systems Beyond First A/B Test
Vijay Pappu – Senior ML Engineering Manager, Personalization Lead at Peloton
2.35pm - 3.05pm
Towards Mass Adoption of Computer Vision Application
Appu Shaji – CEO & Co Founder at Mobius Labs
Appu will discuss how next-gen computer vision has removed boundaries to adoption and is now readily available to everyone, including non-technical users.
During this talk you can see real life use cases of companies that have built AI-powered business applications without a staff of AI experts, complicated integrations or the risk of having their data exposed.
3.10pm - 3.40pm
Using Data to Optimize the Content Acquisition Lifecycle
Tim Yoo – Sr. Director, Head of Analytics at Roku Inc
The content acquisition lifecycle is complex and multi-faceted. Streaming services that strive to be both a competitive service and a meaningful source of entertainment to consumers need to efficiently value content as well as understand the complex relationship between value to the company and to consumers. Discover how analytics and modeling can be used to navigate through the various stages of the content acquisition lifecycle.
3.45pm - 4.15pm
ESG as the signal in the noise: Using NLP and verified data assets to create a holistic measure of company resiliency
Rochelle March – Head of ESG Innovation and Analytics at Dun & Bradstreet
Global changes have impacted countries and companies everywhere. From climate change, the Covid-19 pandemic, resource constraints and demographic fluctuations challenge the stability of even the longest-established enterprises. Traditional financial data is not enough today to provide a clear enough picture on sourcing, investment and insurance decisions. ESG data and metrics can serve as valued information and tools for competitive advantage. For Dun & Bradstreet, this means extending its efforts around business transparency to generate ESG intel that can help customers, investors and other stakeholders identify which companies are actively moving towards a different, and hopefully, more sustainable future. This presentation will showcase analysis that explores the relationship between ESG and financial performance, and will provide a deep dive into the NLP machine learning and analytical techniques used to create Dun & Bradstreet’s new ESG Rankings dataset and model.
4.20pm - 4.50pm
Recommendation Strategies for engaging with audiences of Conde Nast brands
Sriram Subramanian – Head of Data Science & Engineering at Condé Nast
Conde Nast is home to many iconic brands with wide ranging and influential content that engages audiences around the world. Audiences find our content via many channels: social, email, organic, and search. In this talk, we outline the machine learning-based recommendation and personalization strategy employed across all of these channels. Key elements of our strategy include: * Social: The most relevant content is promoted on social media *Email: Newsletters are personalized with content of interest to the recipient *On-site: Consumer experiences are enhanced by recommendations based on not only reader interest but also the context *Search: Editorial topic recommendations based on search trends. We walk through several of these use-cases and their positive impact on audience engagement.
A SAMPLE OF TOPICS COVERED
Putting ML Apps Into Production
Personalization At Scale With AI
Cloud Automation And Machine Learning
Natural Language Processing & Deep Learning
Engineering For Data Science
Scaling ML Production
Improving Data Quality
Machine Learning Best Practices
Data Strategy And Governance
Data Ethics And Bias
Designing ML Pipelines Efficiently
Deep Neural Networks
Image Recognition With ML
Using Large Scale Data Sets
Data Science Teams: Managing, Building, Collaboration
Integrating Open Source Tools Into Your Workflows
Content Personalization And Monetization
Data And AI For Emerging Platforms
Data Governance And More!