UnionML – a Python framework for building ML microservices
Hi HN! Niels here. I'm the creator of UnionML (https://github.com/unionai-oss/unionml), a Python MLOps framework that removes the boilerplate and friction associated with building and deploying machine learning systems to production. I've been training and deploying models for almost a decade now, and one pain-point I've consistently had is managing the complexity of building and maintaining an ML stack that works for the entire model development lifecycle - from prototyping to production. UnionML is built on top of Flyte (https://www.flyte.org) and exposes a…
In plain words
UnionML is a Python framework for building machine learning microservices that simplifies the process of taking models from development to production. Built on top of Flyte, it uses decorators to define ML application components and automatically packages them into microservices for training, batch prediction, and online prediction. The framework targets data scientists and ML engineers who want to reduce complexity and boilerplate when managing the full model lifecycle.
written from the facts on this page · September 2026
From the sources
In the maker’s words, at launch
Hi HN! Niels here. I'm the creator of UnionML (https://github.com/unionai-oss/unionml), a Python MLOps framework that removes the boilerplate and friction associated with building and deploying machine learning systems to production. I've been training and deploying models for almost a decade now, and one pain-point I've consistently had is managing the complexity of building and maintaining an ML stack that works for the entire model development lifecycle - from prototyping to production. UnionML is built on top of Flyte (https://www.flyte.org) and exposes a functional interface for defining the building blocks of your ML application via decorators -- think Flask or FastAPI method endpoints -- and UnionML takes care of bundling them into microservices for different use cases such as: - model training - batch prediction - online prediction - (more coming soon!) This project aims to unify the rich ecosystem of data, ML, and MLOps tools that have emerged over the last decade or so (e.g. MLFlow, Sagemaker, Spark, etc.) to provide a nice UX for model developers, in both individual and team settings.
More ai this month
the category →
I trained a 125M-parameter transformer to autocomplete piano performances in real time (~108 notes/sec on an iPhone 15). The idea is basically GitHub Copilot or Tabnine, except instead of prompting it with code, you prompt it by playing a few notes on a MIDI piano. The model then continues what you played, entirely on-device. The app is free if anyone wants to try it. Happy to answer questions about the model, training, Core ML, or the many things that didn't work.
AI · 17d ago · simedw.com
Astute▲585Automate your B2B brand going viral, with new media creators
AI · 18d ago · company-app.joinastute.com
Hey HN, Henry from Cactus here! We previously released Cactus Needle, a 14MB agentic LLM for tool call, device use, and structured extraction for phones, wearables, smart homes, small robots and microcontrollers. We got really great feedback here, and have now incorporated the suggestions to release Needle 2. The whole model is a single 14MB binary that runs a full session in 28MB of RAM; 45m parameters at 2bit compression. Needle hits 500 tokens/sec decode speed on a Raspberry Pi 5, sits between 400-1,500 tokens/sec on VR devices like Meta Quest 3S and Apple Vision Pro, and ranges…
AI · 26d ago · cactuscompute.com
Launched alongside, June 2022
the whole month →- AF
I've seen some interests in (simulation) video games here on HN so I thought I'd share a short version of our story. More than 6 years ago, me and my friend from university were playing around with an idea of making a game we always wanted to play. We worked on it on weekends but the progress was quite slow, especially due to so many dead ends and wasted effort. Eventually however, we solidified our direction and decided to take the risk to resign from our well paid SWE jobs and work on it full time. It took more than a year but yesterday we have finally released it on Steam:…
Life & fun · 2022
- AC
Life & fun · 2022 · benoitessiambre.com
- IS
Life & fun · 2022 · github.com
