Too Many Ideas, Too Little Data

Explore the challenges and solutions of building data pipelines from scratch in this insightful conference talk from ML Conference 2018. Learn how data scientists and product owners can overcome the cold start problem when faced with insufficient data to answer new questions or build solutions. Discover strategies for leveraging open data sources, web-scraped information, and collected data such as logs and sensor data to create robust data pipelines. Gain valuable insights from an insurtech startup's experience in developing data products that address customer needs, even when starting with "zero data." Understand the importance of focusing on customer problems when collecting new data and how to utilize various data sources to enable innovative machine learning solutions.

Too many ideas, too little data | Markus Nutz & Thomas Pawlitzki | ML Conference 2018