These tech giants have one thing in common: they are the best in using data to help users find their favourite show to watch, book the perfect travel destination, find a ride or their favourite music.
Netflix, Airbnb, Uber and Spotify have totally disrupted the markets they operate because they’ve found smarter, faster, and better ways of providing exceptional customer experience by relying on data analytics, AI and ML. And they are all coming to the Data Innovation Summit to present their data workings that contribute to their massive success.
The data engineering and data infrastructure behind your Netflix recommendations
Dao Mi, Senior Data Engineer at Netflix, is presenting on How to Data-Ops at Netflix Studios at the Data Engineering Stage on August 20th.
In an industry where most records are still kept on paper, it’s a unique challenge to both collect data and surface insights to key stakeholders within the company, states Dao. In his session, he will endeavour to showcase how this problem is tackled at Netflix scale. Dao will also reveal novel use cases that call for creative data collection, what scale means in the entertainment industry, how it’s different from a typical tech company, and how to produce operational insights on demand.
Later the same day, Romain Cledat, Senior Software Engineer, Machine Learning Infrastructure at Netflix, will be talking about Using Metaflow to host models for real-time inference at the Accelerate Stage.
Romain is working in the Machine Learning Infrastructure team at Netflix as part of the Metaflow project which aims to make it easy to build and manage real-life data-science projects. His presentation will focus on how the Open Source software Metaflow can be used to deploy and host models for real-time inference. Romain will also discuss the advantages of immutable deployments with lineage information.
The following day on August 21st on the Accelerate Stage, Jason Ge, Senior Software Engineer at Netflix, will deliver his presentation on Using Metaflow in R for scalable and reproducible data science.
Jason is also part of the Netflix Machine Learning Infrastructure team building a human-centric machine learning platform Metaflow. At Netflix, Jason has been helping data scientists improve their productivity across different use cases leveraging Metaflow: a human-centric and developer-friendly infrastructure toolkit.
How Airbnb provides recommendations two-sided travel marketplace
Coming from the core Search and Relevance team at Airbnb, we have Liang Wu, Machine Learning Data Scientist at Airbnb, who will be taking the Accelerate stage to talk about Recommendation in a Two-Sided Travel Marketplace.
A two-sided travel marketplace is an E-Commerce platform where users can both host tours or activities and book them as a guest. A travel recommender system needs to both understand characteristics of its inventories and to know the preferences of each individual guest. In his presentation, Liang presents Airbnb’s efforts on building a recommender system for Airbnb Experiences, a two-sided online marketplace for tours and activities.
Additionally, Liang will provide more in-depth insight into:
- Knowledge Graph Expansion for Tourism Applications: Classic knowledge graph research focuses on building concepts that are more generic, and domain-specific knowledge graph research rarely discusses topics related to travel and tours. By contrast, we precisely focus on the travel domain that involves tightly with locations and extend the generic terms into more location-specific concepts.
- Recommendation with Limited Data Availability: Additional information we can utilize is user profiles, such as travel destination and user origin. Due to lack of user-item interaction data, we find directly using the categorical information easily leads to overfitting. Instead, we propose a novel method of dealing with categorical features.
How to lead data science teams: Behind the scenes at Uber
Juan Manuel Contreras, Data Science Manager at Uber, is coming to the Analytics and Visualisation Stage at the Data Innovation Summit with his Data Science Manager experience on How to lead data science teams: The 3 D’s of data science leadership.
As Juan says, despite a growing demand for data science managers and the unique difficulties of managing data science teams, few resources exist to support aspiring and practising data science leaders. In his presentation, he will demonstrate a framework that defines data science management and outlines three areas of competence needed to succeed as a data science leader: diplomacy, diagnosis, and development.
Spotify’s way of making data dance with analytics and demystifying AI
Completing this tech giants series, we have Jacob Olsufka, Visual Analytics Engineer at Spotify, will present his data visualisation expertise in his session Spotify: Making data dance with Analytics on August 20th at the Analytics and Visualisation Stage.
Jacob will divulge how they turn massive amounts of data into insights that help inform product decisions across the company. He will also outline the challenges they faced with moving into a more data-driven approach and automation and how they tackled them.
The next day, Ekaterina Garbaruk Monnot, Product Manager and Fredrik Schmidt, Engineering Manager at Spotify will convey their story about Demystifying AI in the organisation at the Data Engineering Stage. They will reveal how Spotify works by levelling up ML and AI knowledge in the organisation.
Moreover, Ekaterina and Fredrik will cover some crucial points for companies starting with ML and AI like what to think about when working with Data in a large organisation with autonomous teams and how to innovate with ML across a diverse ecosystem of software products.
Ekaterina and Fredrik will relate the importance of integrating the knowledge about AI and ML in every domain of the organisation and benefit from it. This can be achieved through educating Managers and non-coding employees about AI and ML and how they could incorporate it in the products.
They will reveal why Integrating good data practices throughout the organisation and with external partners is no longer a “technical solution”, but it is a business strategy. While, at the same time why although ML itself is a solution, it is not fit to solve every problem.
Finally, Ekaterina and Fredrik will share practical examples of good and bad data practices.