Case Study
Nordic Public Transit Modernization
Real-time event-driven platform for national-scale ticket vending machine infrastructure.
Disclosure
This project was delivered by SnowIT's founder during a prior senior consulting engagement (2022–2024) through a Norwegian consultancy partner. SnowIT d.o.o. was founded in 2022 and does not yet carry independent client delivery history at scale. This case study represents the founder's prior work, included to demonstrate domain expertise and technical capability.
Client identity is disclosable under NDA on inquiry.
The Challenge
Legacy ticketing at national scale
A national public transit operator managing 60+ regional transport providers needed to modernize legacy ticket vending machine (TVM) infrastructure serving millions of daily transactions.
Existing systems suffered from fragmented backend architecture, batch-only processing, no centralized observability, and aging Java monoliths with limited test coverage.
Project requirements
- Integrate real-time data from 1000+ ticket vending machines
- Support multi-modal journey planning and fare calculation
- Horizontal scalability for peak commute demand
- Operator-facing dashboards for fleet management and revenue reporting
The Solution
Event-driven microservices platform
Designed and implemented an event-driven microservices platform using Kafka and KSQLDB for real-time stream processing. Replaced synchronous API calls with reactive patterns, enabling millisecond-latency ticket validation and fare calculation.
Key technical decisions
- Kafka Streams topology — modular stream processors for ticket validation, fare calculation, inventory, and audit logging
- API security layer — Apigee API Management, OAuth2/OIDC, rate limiting for external partner integrations
- Kubernetes orchestration on GCP — auto-scaling, blue-green deployments, centralized logging (Grafana/Prometheus)
- Backend modernization — Kotlin/Spring Boot microservices replacing Java monoliths, reactive patterns (Spring WebFlux)
Delivery approach
- Strangler pattern migration — incremental replacement maintaining 99.9% uptime SLA
- Cross-functional collaboration with product, UX, and DevOps teams
- Comprehensive documentation — ADRs, OpenAPI specifications, operational runbooks
Outcomes
Results delivered
1000+
Ticket machines connected
<100ms
Event processing latency
99.9%
Uptime maintained during migration
Millions
Daily events processed
Business impact
- Improved passenger experience — faster ticketing, real-time journey updates, seamless multi-modal planning
- Operational efficiency — centralized fleet monitoring, automated incident alerting
- Revenue assurance — audit-grade transaction logging, fraud detection patterns
Technology
Stack delivered
Backend
API & Security
Cloud & DevOps
Observability
Frontend
Is This Relevant to You?
Similar challenges we can address
This experience applies directly if you are facing:
- Legacy transit or ticketing infrastructure modernization
- Real-time IoT data processing at scale (millions of events/day)
- Event-driven architecture migration (batch → streaming)
- Cloud-native platform design (AWS/Azure/GCP)
- Multi-tenant SaaS for transport operators
Start a Conversation
Discuss your project
We respond within one business day. Initial consultations are confidential and obligation-free.