Backend engineer in search

At Kleinanzeigen, search is the beating heart of how millions of users discover listings every day. We're evolving our search platform — moving to Vespa, expanding semantic and vector search capabilities, and building ranking models grounded in real buyer and seller signals. We're looking for a Senior Search Engineer who bridges the gap between search infrastructure and data science: someone who can both architect a high-performance retrieval pipeline and run rigorous offline experiments to validate relevance improvements.

• Design and implement search features end-to-end: from indexing pipeline changes and schema design to query rewriting, ranking, and autocomplete.
• Lead data-driven technical initiatives — write RFCs grounded in query log and user behaviour analysis, then own delivery from design to production.
• Champion semantic and vector search: design embedding-based retrieval components, integrate LLMs for query understanding or evaluation, and measure real impact via A/B experiments.
• Build and maintain offline evaluation pipelines (relevance judgment, NDCG, MRR) to accelerate the feedback loop between experimentation and deployment.
• Improve indexing reliability and throughput — reduce re-ingestion times, ensure consistency, and tighten SLAs.
• Contribute to the Vespa migration: design index schemas, relevance profiles, and deploy ML ranking models within the platform.
• Collaborate cross-functionally with Data Science, Product, and other engineering teams; represent the team at technical forums and occasionally at external conferences.

• 5+ years of hands-on experience in search engineering — production-scale deployments using Elasticsearch, Solr, Vespa, or similar.
• Strong data science fundamentals: experimental design (A/B testing, interleaving), relevance evaluation, and statistical analysis with Python (Pandas, NumPy, Jupyter).
• Solid software engineering skills in at least one JVM language (Kotlin, Java, or Scala) plus Python.
• Deep familiarity with vector/semantic search: dense retrieval, approximate nearest neighbour (ANN), embedding fine-tuning or prompting for query/document understanding.
• Track record of authoring and leading technical proposals based on data analysis — not just implementing specs written by others.
• Будет плюсом: Experience with Vespa or migrating from Solr/Elasticsearch to Vespa; Familiarity with LLM-based approaches to data labelling, evaluation set generation, or query augmentation; Experience with big data tooling (Spark, Trino, DuckDB) for large-scale offline analysis; AWS cloud experience (S3, DynamoDB, Lambda, Kubernetes); Conference talks, published work, or open-source contributions in the search/IR space

• 28 business days of paid annual holidays
• 20 days “work from anywhere”
• Flexible working time (hybrid)
• Paid additional parental leave for both parents
• Extra paid days off for special occasions (e.g. your wedding or a move)
• Private health insurance
• Gym Pass (may cover up to 70% of gym membership fees)
• 40€ per month as a remote working allowance.
• Personal development budget
• Learning opportunities to grow your skills and career
• A unique position in an international and dynamic environment
• A flat, collaborative culture where your feedback actively shapes our processes
• Competitive compensation package (Base salary + annual bonus)
• Flexible Remuneration Program (tax-efficient options for transport, restaurant vouchers and childcare)
• Corporate benefits and exclusive discounts
• An attractive referral program