Go Microservices for ML at the Edge
summary
A Go microservice pipeline ingests microscope images on Windows OEM devices, sends jobs over MQTT to a Linux control plane for ML inference, and returns outputs — enabling virtual staining that preserves biospecimens. Implementation uses EdgeX Foundry, Redis job repo, Docker on the control plane, and packages like waitforit, testcontainers-go, httpexpect and go-i18n for Chinese UI. Deployment moves from go-msi to ZIP+PowerShell across four test machines, survives firewall polling and intermittent timing/idempotency issues, and proves the pipeline works end-to-end.
AI-written summary · about
chapters (7)
- 0:00Go microservices intro
- 5:44Healthcare ML pipeline architecture
- 13:52Microservice coordination at edge
- 19:35Deployment on Windows/Linux
- 25:39Testing ML pipeline with Go
- 32:22Go microservice i18n challenges
- 40:08Summary
related talks
43mAI/ML Data Pipeline Processing with Go Microservices Based...
21mHarnessing Microservices for Scalable, Fault-Tolerant ML Systems
39mSustainably Automate Manual Cell-Staining with Innovative Open-Sour...
35mEvery ML Model Deserves To Be A Full Micro-service
31mEnabling MEC platforms in Healthcare & Smart Hospitals
29mTeam Teardown: MLOps in the Cloud for IoT
38mMicroservices in Go using Micro
45mAIoT Ops - Using Kubernetes and ML Ops to Build Edge ML Applications