Lead Applied Scientist – AI Search, Brand Intelligence

This is a fully remote position, open to applicants in Europe.

📋 Description

• Analyze and reverse-engineer the mechanisms through which AI search and generative platforms recommend and describe brands.

• Develop the measurement framework to assess the business value of AI visibility.

• Lead the research agenda from hypothesis formulation through experimentation, insights, and the delivery of product capabilities.

• Design and conduct experiments, establishing a repeatable experimental pipeline.

• Set the technical direction, prioritize hypotheses, review methodologies, and oversee the daily operations of a small team.

• Execute the initial 3-4 experiments and identify potential product features in collaboration with Product.

• Create sampling techniques for AI responses across various topics, intents, and regions, addressing issues of variance, bias, and representativeness.

• Establish evaluation frameworks for LLM outputs, prompts, and model behaviors.

• Utilize LLMs through batch APIs as feature extractors and evaluators, encompassing prompt assessment, cost-effective batch inference, and agentic/tool-utilizing pipelines.

• Maintain ownership of standards for model evaluation and validation.

• Develop models for predictive analysis, ranking, classification, clustering, and representation learning.

• Analyze SERP, backlink, content, audit, GA, and GSC data within ClickHouse or comparable columnar OLAP systems at a scale of hundreds of millions of rows.

• Generate and prioritize hypotheses, manage a backlog, and oversee the execution cycle.

• Mentor ML Engineers and engage in the hiring process.

• Transform findings into product features, published research, conference presentations, and GEO best practices in collaboration with Product, Engineering, Marketing, and Leadership.

⛳️ Requirements

• 5-6+ years of experience in Data Science or Applied Science.

• Proven track record of guiding ambiguous research problems from hypothesis through to actionable decisions.

• Extensive, hands-on knowledge in ranking and information retrieval, including learning to rank, NDCG/precision@K, LambdaRank/LambdaMART.

• Strong and up-to-date expertise in GenAI/LLM, particularly in prompting and evaluating commercial APIs and open-weight models.

• Experience with LLM-as-judge or agent evaluation.

• Understanding of how sampling, context, and model behavior influence model outputs related to brands.

• Strong background in model evaluation and validation, including leakage detection, temporal/out-of-time validation, and business-outcome metrics.

• Capability and willingness to guide the daily work of a small team (1-2 members).

• Proficient in English (B2+).

• Familiarity with gradient boosting using LightGBM/CatBoost is a plus.

• Experience with embeddings and representation learning using sentence-transformers/faiss is advantageous.

• Ability to analyze large-scale data directly, such as ClickHouse, is beneficial.

• Interest in the SEO/search domain is desirable.

• Knowledge of AI-search measurement concepts is a plus.

• Experience with MLflow or similar experiment tracking tools is advantageous.

• Publications or speaking engagements are a plus.

🏝️ Benefits

• Direct, ticket-free access to SE Ranking's SERP, backlinks, content, audits, GA/GSC datasets, along with Planable's social media data.

• Dedicated support for data and analytics engineering.

• Allocated budget for LLM API to facilitate systematic response sampling (OpenAI, Google, Anthropic, Perplexity).

• Opportunities to present at conferences, publish research, and outline company best practices.

• Empowerment to initiate experiments without requiring prior approval.