How SepLo Rebuilt Its Backend for a Scalable, Secure Launch on AWS
The Client
SepLo is a multi-portal compliance platform for last-mile delivery logistics, serving drivers, fleet operators, enterprise clients, and administrators through dedicated portals. The platform handles compliance management for couriers, including document verification, scheduling, and payroll tracking. SepLo operates in the German and Austrian markets and is building toward a multi-partner, multi-country scale.
- Industry:Transportation, Logistics, Supply Chain and Storage
- Company Size:1-10
- Country:Germany
Challenges
- Disconnected frontend and backend – The API Gateway integration wasn’t returning data to the frontend, blocking core functionality across all portals.
- Fragmented authentication – Authorization split between Supabase and Amazon Cognito caused broken login flows, including failed password reset and invitation emails.
- Non-operational core features – Document scanning and scheduling tools were not yet functioning ahead of launch.
- Scaling constraints in the backend architecture – With plans to grow from roughly 17 to 100–200 functions, the Lambda-only setup raised questions about long-term maintainability.
- Limited operational visibility – no monitoring, alerting, or centralized security controls were in place.
- Unstable frontend hosting – the externally hosted frontend lacked the reliability needed for a production environment.
As SepLo approached its production launch across the German and Austrian markets, these gaps made it difficult to guarantee a consistent experience across the driver, fleet operator, enterprise, and admin portals.
Solutions
Migrated backend and frontend to Amazon ECS Fargate
Replaced the Lambda-heavy backend and externally hosted frontend with a consistent container runtime built to support growth toward 100–200 services.
Added CloudFront and AWS WAF for the frontend
Improved delivery performance while adding a layer of web application security.
Centralized secrets in AWS Secrets Manager and SSM Parameter Store
Standardized how credentials and configuration are managed across environments.
Introduced CloudWatch
Monitoring, alarms, and GuardDuty Runtime Monitoring, gave the team visibility into system health and runtime activity.
Layered in security hardening
Including IAM password policy, S3 HTTPS enforcement, access and WAF logging, Amazon Macie, IAM Access Analyzer, and KMS-encrypted log groups.
Built CI/CD pipelines via Bitbucket
Automated build, deploy, and rollback to reduce manual deployment risk.
Delivered full infrastructure as code with a handover runbook
Gave SepLo a reproducible environment they can extend independently.
AWS Services Used
- Amazon ECS (Fargate)
- Amazon ECR
- Amazon VPC
- Amazon CloudFront
- AWS WAF
- Amazon CloudWatch
- AWS Secrets Manager
- AWS Systems Manager (Parameter Store)
- Amazon GuardDuty
- AWS IAM Access Analyzer
- Amazon Macie
- AWS KMS
Results
The infrastructure modernization gave SepLo a stable, secure foundation ahead of its production launch, with the visibility and automation needed to scale confidently across new markets.
Delivered a production-ready platformOn schedule, meeting the customer's launch expectations
- Gained real-time visibilityInto system health and runtime activity for the first time
- Re-architected the backendTo scale toward 100–200 services, up from an initial ~17 Lambda functions
- Cut Manual Deployment RiskWith fully automated CI/CD pipelines
- Strengthened Security PostureWith layered controls across IAM, encryption, and access monitoring
- Moved directly into a Cloudvisor Managed Service trialImmediately after go-live
