What challenges did CytoReason face before AI implementation?

Rapid growth in disease-model and patient data created storage and processing bottlenecks. SQL servers could not hold all data, causing slow query runtimes and delaying insight generation. The company needed to meet stricter deadlines as client projects and modeling workloads expanded.

What was the impact of the AI solution?

Reduced query time from about two minutes to 10 seconds. Enabled faster iteration for data scientists and engineers working on disease models. Helped CytoReason meet strict deadlines on certain projects. Supported growth to more than 2.5 PB of disease-model data and over 2,000 BI reports. Contributed to scalable processing for major pharma clients such as Sanofi and Pfizer.

Source

GCP

CytoReason uses BigQuery and GKE to accelerate AI disease modeling for clinical trials (Vertex AI/Gemini not specified)

Use case typeAI development platformUpdated Jun 13, 2026

CytoReason is an Israeli biotech and data platform that creates AI-based computational disease models using public and proprietary data. The company maps human diseases tissue by tissue and cell by cell to help pharma customers shorten clinical trials and reduce drug development costs. CytoReason moved PostgreSQL databases and analytics workloads to BigQuery to store and query very large datasets at speed. It also uses Google Kubernetes Engine for autoscaling and high-performance computing, and worked with WideOps to optimize Kubernetes infrastructure costs. The article says CytoReason built its own high-performance computing solution internally on GKE and uses Cloud Storage plus billing tools for cost optimization.

Organization: CytoReason
Industry: Pharma
Location: Israel
Published: June 2026

Reported outcomes

10 seconds

timeTime & speed

Strategic outcomes

New product / capabilityBuilt an internal high-performance computing solutionSpeed & agilityEnabled faster disease-model iterationSpeed & agilityMet strict project deadlinesScale & capacitySupported larger-scale data processing

Primary read

Use case focus

Showing 3 of 3

1Data Platform Modernization
2Analytics Acceleration
3MLOps / AI Infrastructure

Rapid growth in disease-model and patient data created storage and processing bottlenecks.
SQL servers could not hold all data, causing slow query runtimes and delaying insight generation.
The company needed to meet stricter deadlines as client projects and modeling workloads expanded.

Migrated PostgreSQL databases to BigQuery for faster storage and querying of large datasets.
Used Google Kubernetes Engine for automated cluster lifecycle management, pod and cluster autoscaling, and infrastructure cost optimization.
Built an internal high-performance computing solution on GKE to handle phases with more disease model generation.
Used Cloud Storage classes and billing export reports to improve cost optimization.
Partnered with WideOps to optimize GKE usage and reduce infrastructure spending.

Technologies

BigQuery Google Kubernetes Engine Cloud Storage

Reduced query time from about two minutes to 10 seconds.
Enabled faster iteration for data scientists and engineers working on disease models.
Helped CytoReason meet strict deadlines on certain projects.
Supported growth to more than 2.5 PB of disease-model data and over 2,000 BI reports.
Contributed to scalable processing for major pharma clients such as Sanofi and Pfizer.

Architecture

CytoReason migrated PostgreSQL data into BigQuery for faster analytics and query performance, and runs an internally built high-performance computing solution on Google Kubernetes Engine. The setup relies on GKE autoscaling and health checks, with Cloud Storage and billing tools used for cost optimization, and WideOps assisted with Kubernetes cost and infrastructure optimization.

Implementation partners1

WideOps

Sources & evidence1

Groundedness: 5/5Type: Customer StoryPublished: Jun 3, 2026Publisher: Google CloudEvidence: PrimaryConfidence: High

Primary source

AI-generated summary. Verify important details with the linked sources before relying on this case.

Explore related AI use cases

Customer evidence