Principal Applied Scientist

About the role

As a Principal Applied Scientist at Parloa, you will define and drive our applied AI strategy across AI agents and Generative AI. This is a highly hands-on and high-impact role at the intersection of research and production.

What you'll do

• Define and lead applied AI initiatives across agent systems, LLM evaluation, and model optimization
• Own ambiguous problem spaces end-to-end: from framing and experimentation to production impact
• Design and implement evaluation & benchmarking frameworks, leveraging and challenging industry standards—and where needed, defining new ones
• Drive innovation in agentic systems, including topics like routing, memory, and context engineering
• Prototype and validate new approaches (e.g., model combinations, fine-tuning strategies, or open-weight models)
• Translate research into production-ready solutions, working closely with engineering teams
• Act as a technical authority and multiplier, mentoring others and shaping best practices
• Establish and lead the Applied Science guild, fostering knowledge sharing and raising the bar across teams

What you'll need

• 12+ years of experience in Applied Machine Learning, Applied Science, or a related field, with a proven track record of delivering ML/AI systems in production
• Experience with agent-based systems or multi-step LLM workflows
• Strong recent experience working with LLMs and Generative AI systems, ideally in production environments
• Demonstrated ability to own and drive ambiguous problem spaces end-to-end—from problem framing to measurable impact
• Deep understanding of evaluation methodologies, benchmarking, and model performance analysis, including human-in-the-loop approaches
• Hands-on technical skills, with the ability to prototype, experiment, and ship solutions in collaboration with engineering teams
• Experience working on production systems, balancing speed, quality, and scalability
• Proven technical leadership and influence, with experience shaping direction, mentoring others, or driving cross-team initiatives

Nice to have

• Familiarity with model routing, fine-tuning, or open-weight models
• Background in human-in-the-loop evaluation or annotation systems
• Contributions to public research, blogs, or open-source projects