Senior AI Data Scientist, Agentic Automation

This is a fully remote position, open to applicants in Belgium, +4 more countries.

📋 Description

• Enhance marketing operations at team.blue by creating agentic systems.

• Collaborate with the Applied AI team on projects related to marketing automation.

• Analyze marketing processes across various brands, markets, and languages.

• Measure time, resource usage, and ROI impact of agentic automation.

• Develop proofs of concept, transition them to production, and assess their impact.

• Explore competitive and pricing monitoring, performance reporting and diagnosis, SEO and AI-answer visibility, content refresh, localization and lifecycle production, paid search and social account management, as well as tracking and consent quality assurance.

• Gather requirements from domain experts and transform requests into efficient processes.

• Design state transitions, triggers, branches, API calls, human review points, and failure handling mechanisms.

• Establish guardrails, including dry-run modes, approval checkpoints, least-privilege API scopes, and documented undo processes.

• Create thresholds for human-in-the-loop reviews and organize review queues.

• Develop evaluations for diagnosis accuracy, source validity, brand voice, calibration, and quality of generated output.

• Validate tracking data and quantify any errors found.

• Integrate agents with webhooks and scheduled triggers using idempotent handlers.

• Optimize model selection, routing, costs, and latency.

• Assess vendor APIs, including rate limits, quotas, data models, and integration costs.

⛳️ Requirements

• Over 7 years of experience in building data and ML systems in the industry.

• Proficiency in classical ML and applied statistics.

• Experience working through both sides of the LLM transition.

• Background as the sole or first data/ML hire or functional owner.

• Proven experience delivering systems authorized to operate on live systems impacting real customers.

• Expertise in Python and machine learning.

• Capability to deliver end-to-end solutions, including clear Python code, current tools, containerization, instrumentation, and deployment.

• Practical experience with multi-step, tool-calling LLM workflows, including orchestration, retries, idempotency, timeouts, and partial-failure recovery.

• Familiarity with state-machine design.

• Experience integrating with third-party APIs, covering authentication flows, rate limits, pagination, sandbox behavior, and schema changes.

• Knowledge in cost and latency engineering, including model routing, caching, and batching strategies.

• Understanding of safety practices for action-taking systems, such as staging modes, approval gates, least-privilege scoping, and rollback procedures.

• Skills in evaluation design for generative and agentic outputs, including LLM-as-judge, golden-transcript regression tests, red-teaming, and calibration.

• Proficiency in process mapping and quantification.

• Technical vendor evaluation capabilities.

• Proof of eligibility to work in the application country.

• Preferred: Master’s or PhD in Computer Science, AI, Machine Learning, or a related discipline.

• Preferred: Experience with PromptOps at scale.

• Preferred: Familiarity with martech, ad-tech, SEO tools/APIs, or customer-facing automation.

• Preferred: Experience evaluating generated output in multiple languages.

🏝️ Benefits

• No relocation packages or visa sponsorships provided.

• Commitment to diversity and inclusion.

• Participation in ESG and sustainability initiatives.