Customer Case Study: Outreach.io - Databricks

Outreach.io

Customer Case Study

Outreach.io

Outreach.io helps sales professionals efficiently and effectively engage prospects to drive more pipeline and close more deals.

Vertical Use Case

  • Improve sales efficiency with machine learning
  • Customer intent classification

Technical Use Case

  • Ingest and ETL
  • Machine Learning

The Challenges

The Outreach Sales Engagement Platform helps sales teams intelligently engage with prospects with ML-powered insights. For example, the platform helps reps prioritize prospecting by categorizing customer emails with a positive sentiment (e.g. request for a meeting) vs. a negative sentiment (e.g. unsubscribe). On a daily basis, Outreach ingests huge amounts of diverse data such as emails, unstructured, streaming and time series data. Processing this data and building machine learning applications was a challenge.

  • Infrastructure Management: Using EMR to manage Apache Spark clusters was overly complicated and resource intensive.
  • Data Engineering Complexity: Setting up and maintaining data pipelines required large amounts of time and resources, pulling machine learning engineers from higher value activities.
  • Disjointed Machine Learning: Struggled to build, train, and deploy ML models in a repeatable and reproducible manner. Often times, labeled data wasn’t updated properly in conjunction with the release of a new version of a model, resulting in inaccurate results, delays to production and significant DevOps overhead.

The Solution

Databricks has provided Outreach.io with a fully managed analytics platform that accelerates AI innovation.

  • Fully Managed Platform: A fully managed cloud platform simplifies operations and delivers superior performance of ETL pipelines at scale.
  • Automated Infrastructure Management: Simplified cluster management with auto-scaling significantly reduced time spent on data engineering.
  • Faster and More Reliable Data Pipelines: Removed the complexities of building ETL pipelines. Able to ensure optimal performance for downstream analytics.
  • Robust Machine Learning Infrastructure: MLflow greatly streamlined their machine learning lifecycle, simplifying model reproducibility and process repeatability.
  • Interactive Workspace: Data scientists can collaborate, share, and track data and insights across various programming languages, fostering an environment of transparency and improving productivity.

The Results

Outreach is enabling sales teams to be more efficient and make smarter decisions with machine learning applications built on Databricks.

  • Faster Time to Market: Able to build reliable ETL pipelines that perform much faster – reduced ETL jobs from weeks to days with Databricks.
  • Improved Data Science Productivity: Collaborative notebooks accelerated data science innovation enabling faster development of new applications for their customers.

Databricks really helps us to expedite model development validation and production so that we can give the intelligence power into the hands of the sales reps.

Yong Liu — Principal Data Scientist, Outreach