About Probabl

Probabl is the official and exclusive industrial operator of scikit-learn, the world's most widely used machine learning library. Born from the open-source ecosystem, we are building a comprehensive suite of tools, starting with Skore. Our mission is to help data teams design, track, and validate models through a unique code-first UX, prioritizing ease of use and flow.

The Role

As a Machine Learning Engineer, you will focus on Skore’s interoperability, ensuring it fits naturally into existing enterprise ecosystems. Your primary goal is to drive adoption by removing friction.

You will bridge the gap between our code-first UX and the broader infrastructure. This involves two distinct challenges: building bridges with enterprise tools (e.g. tracking systems, data platforms) and simplifying access to remote compute substrates.

Responsibilities

• Ecosystem Integration: Explore, design, and build integrations with industry-standard tools. Primary candidates for evaluation include MLflow, Databricks, and Snowflake.

• Compute Abstractions: Build remote compute capabilities that allow users to dispatch training or inference workloads to cloud substrates (e.g., AWS, Modal) with local-like simplicity.

• API Design: Craft intuitive Python APIs and abstractions that make complex ML workflows (tracking, artifacts, remote execution) feel simple and consistent.

• Open Source: Actively contribute to Skore Lib, reviewing issues, guiding contributors, and ensuring the open-source core remains high-quality.

• Reliability: Apply continuous integration and testing practices to ensure reliability across diverse environments (local, CI, Hub).

Technical Environment

• Languages: Python (Advanced).

• Libraries: scikit-learn, pandas, numpy, PyTorch.

• Core Stack: FastAPI, PostgreSQL, S3 (Skore architecture).

Requirements

• Experience Level: 3+ years in software development for ML workflows or scientific computing.

• Engineering focus: You are a software engineer building tools for data scientists, not just a data scientist training models. You care about how the plumbing works.

• Ecosystem Knowledge: Strong familiarity with the modern ML stack (e.g. MLflow, Databricks, or similar lifecycle tools) and how to interface with them.

• Compute: Experience programmatically managing compute resources (cloud instances, batch jobs).

• API Design: An obsession with Developer Experience (DX)—you understand why scikit-learn’s API is successful and strive for that level of clarity.

Preferred Qualifications

• Experience maintaining open-source ML libraries.

• Deep understanding of reproducibility constraints in ML pipelines.

Working Conditions

• Location: Paris, Saclay

• Presence: Office-based or Hybrid

• Team: Tech & Product Department / Applications Engineering Team

• Supervisor: VP of Applications Engineering

• Stakeholder Interaction: Regular collaboration with Product Management, Labs, and Engineering.

• Cultural Fit: Should align with the company's values of innovation, agility, and user-centricity, thriving in a fast-paced startup environment.