How a major South American bank accelerated cloud migration with LocalStack

The challenge: Scaling developer workflows for cloud transformation

One core platform initiative involved several hundred engineers building microservices in a monorepo with multiple data stores and AWS-integrated components (databases, messaging/eventing, and other dependencies)

As the organization planned to scale from hundreds to thousands of developers, platform leadership needed to formalize local AWS emulation as a consistent capability: improve support and coverage, streamline CI pipelines, and provide a predictable developer experience across teams.

Risks if not addressed

• Difficulty scaling consistent development practices to thousands of engineers
• Continued CI pipeline flakiness and slower feedback cycles
• Increased likelihood of misconfigurations and security vulnerabilities from ad hoc local setups
• Potential headwinds to achieving the organization’s cloud migration target
• Ballooning AWS spend

What mattered in the decision

To support standardization at scale, the organization prioritized a solution that could:

• Standardize local AWS emulation with predictable behavior across teams
• Integrate cleanly with existing Docker-based workflows and CI pipelines
• Support a phased rollout from an initial pilot to broader adoption
• Provide enterprise-grade foundations (support, coverage, and repeatable workflow patterns)

The solution: Enterprise-scale local AWS emulation platform

LocalStack provided a standardized platform for local AWS service emulation, integrated into Docker-driven developer workflows and CI.

This formalized existing usage while improving consistency and enabling teams to emulate AWS services locally as part of day-to-day development and testing.

Primary capabilities and services supported

• Data stores: DynamoDB and RDS emulation
• Messaging/eventing: SQS and Kafka adjacency testing
• Containers: ECS/EKS testing
• Serverless: Lambda (including layers)
• Storage: S3 testing
• Multi-service orchestration for full-stack microservice development

Implementation: How the rollout was designed to scale

The organization planned a phased approach:

• A time-boxed proof of concept with ~100 engineers on the initial platform initiative
• Purchase for 2,000 engineers upon successful POC completion
• Long-term goal to expand to 15,000+ cloud engineers after validation

Key integrations included:

• Docker Compose + LocalStack for microservice virtualization in a monorepo
• Pre-seeded environments for faster “warm” dev stacks (e.g., database/environment warm-up)
• GitHub Actions CI integration for consistent testing workflows
• Secure access patterns (e.g., SSO-enabled environments) and approved internal distribution methods

Results and benefits

The organization outlined expected and early observed outcomes from formalizing local AWS emulation:

• Developer velocity and pipeline performance: Improved feedback cycles by reducing reliance on real AWS environments for iterative development and service interaction.
• Standardization across engineering: More uniform workflows across teams using Docker + LocalStack, reducing ad hoc setup risk and improving consistency.
• Quality and reliability: Increased test accuracy by virtualizing full-stack components locally and using isolated test environments with no production exposure.
• Cloud migration enablement: Earlier validation in developer loops supports broader cloud transformation goals and reduces friction from limits, cost, or shared environments.

‍