Most machine learning systems stop learning the moment they ship.
We build the Continuous Learning System (CLS), continuous AI: an architecture that learns during live operation.
It ingests streaming data through a mesh of distributed actors and adapts to structural change without a retraining cycle, running entirely on ordinary CPUs.
The CLS runs today in ongoing projects with partners across several industries and use cases, and we are funded.
We are a small, senior team.
This role exists because the system works and the world around it now needs to be built.
Tasks Product.
Build and own our customer-facing surface: dashboards, APIs, account and access management, the interfaces through which customers meet the CLS.
Data platform.
Pipelines on Dagster, time-series storage in QuestDB, and ingestion from the real world in all its mess: industrial CSV exports, biosignal formats such as EEG, EMG, and IMU, and public APIs of famously variable reliability, ENTSO-E being the canonical example.
Live operations.
Build the internal tools that let a team of six move like a team of twenty.
Growth path.
As you learn the system, work your way into the CLS core itself: the distributed runtime and the engineering of a system that never stops learning.
We will actively support that path.
An outstanding recent MSc with substantial real project work also qualifies.
Production experience in at least one modern stack.
Our codebase is Python; if yours was Java, Kotlin, or TypeScript, we expect you to become excellent in Python quickly, and we have seen strong engineers do exactly that.
Comfort with Linux, Docker, and cloud infrastructure.
Familiarity with distributed systems and streaming data, or the appetite to build it fast.
We run an actor mesh on Ray, and hands-on experien.