CS-06 — Backend & integrations · Cloud architecture · Automation & AI
.NET microservices, serverless AWS and ML pipelines
Development and maintenance of C#/.NET microservices and Lambda functions across several areas of a pricing platform, AWS infrastructure support in several regions, integration of ML pipelines with SageMaker and Step Functions, and delivery with Azure DevOps.
- My role
- Software / Data Engineer contributing to several areas of the platform (Wizeline); not the sole owner of any service.
- Period
- Not documented
- Disclosure
- Anonymised details
- Technologies
- C#.NETREST APIsAWS LambdaAmazon API GatewayPostgreSQLAmazon RDSAmazon AuroraSnowflakeAmazon SQSAmazon SNSAmazon EventBridgeAWS Step FunctionsAmazon SageMakerAmazon CloudWatchAWS IAMAWS Systems ManagerAWS Secrets ManagerAmazon MWAAAzure DevOps
- Last reviewed
- 13 November 2026
Context
The platform is made up of microservices and serverless functions for pricing, pricing workflows, strategies, margins, volumes, reporting, station management, control and configuration, entitlements and notifications, on PostgreSQL, Snowflake and AWS. It also includes ML workflows for market-based price proposals.
Challenges
- 01Understanding existing backend architectures and application bugs originating in the data.
- 02Communication between services and distributed messaging.
- 03Multi-environment and multi-region configuration, IAM permissions and secrets.
- 04Execution dependencies and rerun procedures in orchestrated ML pipelines.
Solution & architecture
Development of APIs and integration logic in C#/.NET with dependency injection and asynchronous messaging; database access optimisation; dependency upgrades; resolution of production incidents with CloudWatch; support for serverless and cross-region deployments; and REST endpoint testing, regression testing and post-deployment validation. In the ML pipelines (SageMaker, Step Functions, Lambda, EventBridge, Snowflake, SSM), investigation of orchestration, authentication and connectivity with Snowflake, execution dependencies, restart and rerun, scheduling and environment configuration.
Key decisions
Not documented yet
Trade-offs
Not documented yet
Results
Results will be published once verified
Lessons learned
Not documented yet
In the ML pipelines, Ander's role is integration, investigation and infrastructure support; he is not the author of the research or the models.
Delivery: Azure DevOps, Git, pull requests, code reviews, release pipelines, promotion between environments (development, test, pre-production, production) and deployment coordination.
Ask about this case study
Questions here are answered with the case study “.NET microservices, serverless AWS and ML pipelines” as context.