AI in Go: how we build ML apps using the pipeline pattern
summary
Describes how T-Bank builds ML applications in Go focusing on the backend—model serving, pipelines, and integration—treating models as black-box components alongside queues and databases. Explains a pipeline pattern implemented in Go for production ML workflows (serving, orchestration, and backend engineering concerns rather than model internals).
AI-written summary · about
chapters (12)
- 0:00Introduction
- 1:53What is ML app
- 3:42Why pipelines in ML
- 6:53Why Go for pipelines
- 7:38Live coding: simple pipeline
- 11:54Adding audio logger
- 14:46Pipeline library
- 17:11How library works
- 20:44Library benefits
- 23:54Production usage
- 25:20Summary
- 26:02Q&A
related talks
40mAI in Go: how we build ML applications using the pipeline pattern
49mMachine Learning Pipelines: What No One Tells You About Building AI
38mAI in Kubernetes: How to Get Started?
33mAI with a DevOps mindset - experimentation, sharing ..
41mInfrastructure for ML Applications
14mSupercharging ML Pipelines with Go
17mContinuous deployment for ML: The new software development lifecycle
27mAI as a Service Scaling AI with Ease and Innovation