What challenges did Guru face before AI implementation?

Conciliate high-volume financial market data and real-time access for customers. Deliver reliable app experiences while handling millions of transactions and high API traffic. Replace a multi-cloud setup with a more efficient and flexible platform.

What was the impact of the AI solution?

~150% increase in production deployment frequency. 1.7 million API requests per month. 139 ms average response/processing time. 0.06% service error rate and 99.9% of users never experienced app errors. 60% reduction in cloud costs and improved productivity.

Source

GCP

Guru case study (Google Cloud Customers) — financial services & insurance

Use case typeCloud migrationUpdated Jun 13, 2026

Guru is a Brazilian fintech for investments that migrated from a multi-cloud setup to Google Cloud to improve cost efficiency, resource optimization, and throughput. The company redesigned its architecture around Cloud Functions as the core execution path, backed by a caching layer on Google Kubernetes Engine, and uses BigQuery for data management. The article notes future and ongoing work to derive more consolidated data views with BigQuery and to explore AI use cases, but the implementation described is primarily cloud infrastructure modernization rather than an AI solution.

Organization: Guru
Industry: Finance
Location: Brazil
Published: April 2023

Reported outcomes

+150%

quantified impactOther quantified impact

0.1%accuracy99.9%accuracy−60%cost

Strategic outcomes

Cost efficiencyImproved cloud cost efficiencyNew product / capabilityModernized cloud architecture for app executionCustomer experience & trustDelivered reliable app experiencesSpeed & agilityIncreased production deployment frequency

Primary read

Use case focus

Showing 3 of 5

1Cloud migration
2Platform modernization
3Serverless applications

Conciliate high-volume financial market data and real-time access for customers.
Deliver reliable app experiences while handling millions of transactions and high API traffic.
Replace a multi-cloud setup with a more efficient and flexible platform.

Migrated fully to Google Cloud.
Used Cloud Functions for most application execution.
Added a caching layer on Google Kubernetes Engine to support containerized apps.
Used BigQuery for database management and data warehousing.

Technologies

Cloud Functions Google Kubernetes Engine BigQuery

~150% increase in production deployment frequency.
1.7 million API requests per month.
139 ms average response/processing time.
0.06% service error rate and 99.9% of users never experienced app errors.
60% reduction in cloud costs and improved productivity.

Architecture

Guru redesigned its platform on Google Cloud with Cloud Functions as the core execution path, a caching layer on Google Kubernetes Engine for containerized apps, and BigQuery for data management/data warehousing.

Sources & evidence1

Groundedness: 4/5Type: Customer StoryPublished: Apr 1, 2023Publisher: 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