Scaling Smart: Legacy Modernization in the European Postal Service

The client’s comprehensive platform helps users manage and track postal shipments via a browser-based interface. Yet behind this interface, the backend systems were burdened by outdated tools, limited automation, and insufficient real-time visibility into performance and failures.

The client’s core business is delivering postal services, including letters and parcels, as well as supplying customers with other post-related services like money transfers and marketing services throughout the country.

Thus, like any other large enterprise, the client could not stop the business operations to revamp the web cloud, modernize outdated tech, and implement strong Information Systems Support (ISS). That’s why they asked us to act in parallel to current operations and according to the latest privacy requirements, avoiding business interruption on the go.

Over 30 engineers in agile teams collaborated across architecture, DevOps, infrastructure, and quality assurance disciplines.

Apart from the need to act in the staging environment to mitigate the downtime risks, the project was characterized by other serious challenges:

• Salesforce’s rigidity limited its infrastructural flexibility, so we had to review existing Salesforce customizations and integrations and explore potential solutions such as events for asynchronous integration, external services, or the use of middleware.
• Frequent misconfigurations occurred across agent setups needed to implement a centralized configuration management system or tool for agent setups.
• Limited access to the production account, that’s why the development was burdened with creating roles and permissions to real-time data access,
• Previous performance tests were manual and reactive, hence, we needed to automate performance testing as part of the continuous integration/continuous delivery (CI/CD) pipeline, e.g., nightly builds, on every significant code change, and add real-time performance monitoring.
• Infrastructure scripts (e.g., Bash) lacked maintainability, so we were forced to refactor existing Bash scripts to improve readability, modularity, and error handling.
• Limited automated scaling options in terms of legislation (data residency laws, access control and auditability, restricted cloud zones, and security certifications).
• High application latency and slow response times were impacting user experience and operational efficiency.

• CI/CD: Jenkins, Groovy, Bash
• Monitoring: Grafana, InfluxDB
• Deployment: Jenkins Pipelines
• Interfaces: Web cloud apps
• Legacy: ISS, Salesforce

The DevOps strategy focused on continuous releases, automation, and using key metrics to ensure timely results while minimizing manual updates in production. Key changes included urgent fixes and refactoring, along with proactive actions to ensure further maintainability and scalability of the system:

• We’ve tackled the infrastructure script maintainability issues by refactoring existing Bash scripts. This effort significantly improved their readability, modularity, and error handling, making our infrastructure management much more robust.
• Manual deployment scripts were moved into structured Jenkins pipelines to ensure automated, repeatable, and traceable deployments. Moreover, we’ve designed renewed CI/CD workflows according to the latest practices for fast, controlled releases.
• We’ve created a complex role-based production access system for performance monitoring according to the current legislation norms in the field of data protection. Moreover, implemented IAM policies and encrypted data at rest and in transit complied with GDPR and internal security guidelines.
• To resolve frequent agent misconfigurations, we’ve introduced a centralized configuration management system. This approach ensured consistent and accurate agent setups across the board.
• We’ve automated scaling and alerting within legal constraints, and for further cost-optimization, we’ve used auto-scaling groups and cost-monitoring tools to minimize infrastructure spend without sacrificing performance. This dynamic scaling ensured the infrastructure can rapidly adjust to demand fluctuations, remaining low-latent and highly available even during peak loads.