ml engineer for real estate search

AVIV Group is a digital real estate technology company that operates recognized European platforms for buying, selling, renting, leasing, and valuing real estate. It combines data, technology, and human expertise to help millions of people find homes.

• Design, train, and ship models for search relevance, ranking, and recommendations across candidate retrieval, re-ranking, and business-rule layers;
• Set the experimental standard through offline evaluation, A/B testing, and evidence-based decisions about model performance;
• Own the architecture of touched systems, including build-versus-adopt decisions, reproducibility, and observability;
• Operate shipped systems through serving, monitoring, SLOs, and participation in the on-call rotation;
• Lead the move toward LLM-augmented search, including query understanding, semantic retrieval, and conversational surfaces;
• Mentor through design reviews and technical writing;
• Work directly with product teams to define optimization targets and challenge measurable but incorrect goals with evidence.

• 8+ Years of engineering experience, including at least 3 years shipping ML systems to production at meaningful scale;
• Hands-on experience designing ranking, retrieval, or recommendation systems across the full funnel;
• Experience with candidate generation, feature engineering, model architecture, and re-ranking;
• Experience diagnosing offline/online divergence and improving a team's evaluation practice;
• Strong engineering fundamentals, including production-quality code, reliable data pipelines, and reproducibility;
• Track record of technical influence beyond personal output;
• Clear communication about uncertainty, including “we don't know yet”;
• Nice to have: LLM-based search features in production, semantic retrieval, embeddings at scale, RAG, query understanding, learning-to-rank, vector search, candidate-generation systems at high scale, marketplace or classifieds domain background, geospatial features in ranking.

• No conditions specified