--- title: "GopherCon 2025: Supercharging ML Pipelines with Go - Vaidehi Thete" url: https://daily.dev/posts/gophercon-2025-supercharging-ml-pipelines-with-go---vaidehi-thete-tg77dzm3q source_url: https://www.youtube.com/watch?v=mLGDT2TxPrU type: video:youtube source: "GopherCon (Gopher Academy)" published: 2026-05-31T07:46:49.807Z updated: 2026-05-31T07:57:57.496Z tags: ["machine-learning", "golang"] reading_time: 14 upvotes: 0 comments: 0 language: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # GopherCon 2025: Supercharging ML Pipelines with Go - Vaidehi Thete **[GopherCon \(Gopher Academy\)](https://daily.dev/sources/gopheracademy)** · 14 min read · 0 upvotes · 0 comments ## Summary Engineers at the New York Times ML platform team share how they used Go to operationalize a Python-based ML ranking model for targeted email campaigns at scale. The core challenge was bridging a Go-based platform with a Python ML model to generate 40M+ user-content recommendations. The solution uses interprocess communication via shared memory (shims), where a Go service packs tensor data into shared memory regions that a Python inference server (NVIDIA Triton) reads directly. Go handles initialization, batching tensors by type, checksum validation, worker pools, and goroutines for throughput. The result: cohort generation time dropped from hours to minutes, with statistically significant gains in user engagement. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: --- Tags: [#machine-learning](https://daily.dev/tags/machine-learning), [#golang](https://daily.dev/tags/golang) [View this post on daily.dev](https://daily.dev/posts/gophercon-2025-supercharging-ml-pipelines-with-go---vaidehi-thete-tg77dzm3q)