About :probabl.

We are the company stewarding scikit-learn, the world's most widely used machine learning library. Spun out of Inria, headquartered in Paris, we exist to keep open-source scientific Python at the heart of how the world builds Data Science — and to make it work for serious enterprise use.

Our product portfolio extends scikit-learn into the full Machine Learning lifecycle:

• Skore — an ML evaluation platform that turns model development into a rigorous, reproducible data science practice

• Skolar — interactive learning for the next generation of data scientists

We work hand-in-hand with the scikit-learn maintainer community, partner with leading enterprise design partners, and operate at the intersection of open-source culture and enterprise ML.

The Role

We're looking for a Product Marketing Manager to own how Skore — and the broader :probabl. portfolio — is positioned, launched, and adopted by data scientists and ML engineering teams worldwide.

This is a high-leverage role at a pivotal moment. The MLOps category is consolidating (Neptune → OpenAI, W&B → CoreWeave), tabular foundation models (Fundamental.tech, Prior Labs’ TabPFN, INRIA’s TabICL) are reshaping the technical landscape, and the rise of agentic coding is creating urgent demand for methodological guardrails — as more data science work runs through AI assistants, teams need ways to keep their workflows rigorous and reproducible. We have a credible, technically grounded answer in Skore — and we need someone to make sure the right teams hear it.

You'll report to the CPO and work closely with Product, Engineering, the CEO, and our CSO Gaël Varoquaux (scikit-learn co-founder).

What you'll do

• Positioning & messaging — Sharpen how Skore stands apart from MLflow, Weights & Biases, Neptune, and emerging agentic ML tooling. Translate technical depth into a story practitioners trust and buyers can champion internally.

• Developer marketing — Drive top-of-funnel through technical content (blog posts, landing pages, comparison guides, GitHub-native proof points), conference presence (PyData, EuroSciPy, Open Source Summit), and the scikit-learn community.

• Launches — Own end-to-end go-to-market for Skore releases and broader portfolio milestones — from narrative and assets to enablement, press, and partner coordination.

• Design Partner Program — Partner with Product and Sales to scale our flagship enterprise motion into a repeatable program with reference customers, case studies, and a clear evidence base.

• Competitive intelligence — Maintain a living view of the MLOps and agentic data science landscape. Equip the team with battle cards, win/loss analysis, and crisp differentiation.

• Demand drivers — Build the marketing motion around regulatory tailwinds and translate them into concrete buyer narratives.

• Voice of the practitioner — Run customer and community research with Product to keep our messaging grounded in what data scientists actually care about.

• Sales, employees & Partners enablement : create material and collaterals that will be used to on-board and allow impactful market interactions for Sales, new employees and partners on Probabl value proposition and products.

• Feed our marketing efforts : feed the marketing activities (campaigns, web site, events, …) with the content and messages that best communicate our products and services descriptions and benefits.

What we're looking for

• 4–7 years in product marketing for developer tools, ML/AI platforms, data infrastructure, or open-source-led B2B software.

• Technical fluency. You can read Python, sit comfortably in a Jupyter notebook, and have an opinion on train/test split vs cross-validation. You don't need to be an ML engineer, but you need to be able to talk credibly with one.

• Open-source instinct. You understand that developer marketing in OSS communities is earned, not bought — through contribution, transparency, and respect for the practitioner.

• Sharp writing. Strong sample portfolio of technical blog posts, launch narratives, or positioning work. You can take a 40-page architecture doc and produce one paragraph that lands.

• Launch operator. You've shipped a handful of meaningful product launches and can show the playbook.

• Bilingual — fluent in English (our primary working language) and comfortable in French.

Nice to have

• Experience marketing to data scientists, ML engineers, or quant teams.

• Familiarity with the scikit-learn / PyData ecosystem.

• Background or interest in MLOps, model evaluation and experimentation.

• Hands-on experience with Claude Code or other coding agents.

• Track record building or contributing to an open-source community.

• A point of view on the agentic data science wave — and where it's hype vs. real.

What we offer

• A seat at the table during the foundational phase of a category-defining AI company.

• Direct collaboration with the scikit-learn founding team and the maintainers shaping the future of the library.

• Real ownership over how :probabl. shows up in the world.

• Competitive compensation.

• Hybrid working from our Paris office, with flexibility for remote work.

• Conference and OSS contribution budget.

• The kind of intellectual seriousness you'd expect from an Inria spin-off — without the bureaucracy.