Head of Data and Privacy Engineering

Join GlobalML in a founding role: own its technical side, contribute to its strategy, and turn real company data into lawful, privacy-safe training data for AI. Full-time, remote, from France or the San Francisco Bay Area. ABOUT GLOBALML GlobalML turns the complete record of companies into cleared training data for AI agents.
• The market. AI labs and AI data partners are starving for real-world datasets. Anthropic reportedly discussed spending over $1 billion on RL environments within a year, and Google won the auction for the Spirit Airlines bankruptcy data estate at $10 million, with Mercor as backup bidder at $7.5 million.
• The data. Code with its full history, the dev workspace (PRs, issues, CI) and the business workspace (email, chat, docs, tickets). The code and its vulnerabilities, the events, the decisions and the timeline stay real. Everything that could identify a person is rewritten and every secret is removed, so buyers receive anonymized data.
• The sources. European liquidations through court-appointed liquidators, US receiverships and assignments, and licensing from operating companies in Europe and the US.
• Where we are. Several repos and company workspaces acquired with clean title and signed export agreements, 100+ repos and workspaces evaluated, and direct relationships with frontier AI labs and AI data partners.
• The founders. Gilles Raymond, CEO, serial entrepreneur (News Republic, sold to Cheetah Mobile, then ByteDance). Peter Mikhalev, COO, co-founder of eTelecare, taken to a Nasdaq IPO. THE ROLE
• Define and build the platform. Identify or build the appropriate tools, skills and technical solutions, and hire the team the roadmap calls for.
• Assess estates. Set how we capture repos and workspaces, and evaluate them before we buy or license: volume, quality, privacy risk and value to buyers.
• Clear repos. Export Git history, PRs and CI. Remove secrets, names, emails and customer data, and keep the code and its vulnerabilities intact. Start with the repos we already hold; more are in the pipeline.
• Clear workspaces. Pseudonymize email, chat, tickets and docs in different languages, and rewrite what identifies people with LLMs while the events, decisions and timeline stay real. Validate the method with a buyer before we build the first paid corpus on it.
• Build compliance in. Make sure raw data is processed only on self-hosted or European-hosted models, and that raw European data never leaves the EU. Set our raw-data retention policy and the anonymization claims it supports. Choose the transfer framework (Data Privacy Framework or standard contractual clauses) for each delivery. Keep the provenance records that help model providers meet their EU AI Act duties.
• Produce the evidence, face buyers. Write the re-identification risk report and processing records buyers rely on, and answer the technical, security and privacy questions of AI labs and startups. YOUR SEAT AT THE STRATEGY TABLE
• Strategic guidelines. Contribute to the company’s direction with the CEO and COO, and initiate strategy adjustments when the data, the buyers or the market call for it.
• Corporate goals. Shape the goals with the founders, and own the technical roadmap that delivers them.
• Technical choices. Lead them: make the call, or bring the founders a clear recommendation.

WHAT WE OFFER
• A founding seat: founder-level equity with 4-year vesting and a 1-year cliff, a salary once funded, and a real role in shaping and executing strategy and corporate goals.
• Data nobody else has: the real, never-public history of companies, with clean title.
• Ownership of the technical architecture and the data and privacy stack, from day one.
• Direct contact with frontier AI labs: you are their technical and privacy counterpart.
•

Location:
full-time and remote, from France or the San Francisco Bay Area. Either way, raw European data must stay in the EU. WHO YOU ARE
• The person who will stop a sale when the data is not safe.
• Obsessed with execution: an entrepreneur’s mindset, focused on shipping with quality.
• A strategic mind: you link technical choices to the business and its goals.
• Highly autonomous and strongly committed, with a will to learn and adjust to business needs.
• Transparent and a clear communicator, at ease internationally, fluent or bilingual in English.
• Five years or more in data or privacy engineering, including time in a small startup. WHAT YOU BRING
• Production code in Python, and Git internals, including history rewriting. TypeScript (the platform we will build) a plus.
• Hands-on removal of personal data and secrets from code and text with NLP or LLMs, in several languages.
• GDPR privacy engineering, and working knowledge of the Data Privacy Framework and EU AI Act, or the drive to master them fast.
• Technical leadership: architecture, build-or-buy, hiring.
• A plus: security, AI training data or evaluation work, French. RL is not required. HOW WE HIRE: FIVE STEPS 1. A 30-minute call with Gilles, CEO 2. A call with Peter, COO 3. A technical evaluation 4. A final call with Gilles 5. An offer GlobalML welcomes applicants of every background. We consider all qualified candidates without regard to race, color, religion, sex, sexual orientation, gender identity, national origin, age, disability or any other protected characteristic. We use your application only for this hiring process and delete it within two years if we don't hire you. To apply, email join@globalml.com with a few lines about you, a link to your LinkedIn, and a link to something you built. More https://claude.ai/artifact/R8kVxiex7jRPHyrZK2u1X5