Business Impact

  • ML Workflow Automation

  • Low Latency for Predictions

  • Rapid Testing & Deployments

  • Scalable Architecture on GCP

Customer Key Facts

  • Location : North America
  • Industry : Insurance

Problem Context

Customer, a leading American commercial property and casualty insurer, had developed a claims model,  but multiple separate modules are required to be integrated seamlessly to provide the expected outcome. They wanted to reduce the complexity in the current architecture and needed a model orchestrator to automate end-to-end machine learning workflows.



  • Integration with multiple modules to streamline end-to-end testing
  • Enabling secret management in the CI pipelines as an upgrade from previous in house methods


Kubernetes Engine
Cloud Composer
Cloud Storage
Apache Airflow


Quantiphi developed a claim modeling orchestrator for real-time and batch processes, ensuring the orchestrator uses a scalable data architecture to ingest data from BigQuery, apply transformations on the data, incorporate multiple modules (claims model, rules engine), and provide an output to BigQuery.


An automated and scalable pipeline orchestration framework to enhance data workflow management, experiment tracking and reproducibility

Looking for similar project?

Let's Talk

Get your digital transformation started

Let's Talk