At John Deere, I architect Go services for equipment telemetry, defining gRPC contracts and service boundaries so teams can ship changes independently. I built Kafka consumers that process roughly 1.5 million equipment events daily, using idempotent handlers and retry policies to protect downstream records.
I moved services to Kubernetes and added Prometheus metrics, Grafana dashboards, and OpenTelemetry traces. These changes shortened production incident investigations from hours to under 40 minutes.
At SysPay, I built Go microservices for payment authorization and processed about 700,000 payment events daily through Kafka. Idempotency checks reduced duplicate transaction handling by 35%, and I used Redis to cache merchant configuration and rate limits.
