SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Salary: USD 250,000 - 320,000 / annual
Cantina Labs is a social AI company building advanced real-time models for AI character creation and storytelling. You will be hired as the first dedicated engineer to own search and recommendations end-to-end for the platform.
In this role, you will own the complete architecture and roadmap for search and recommendations across all discovery surfaces: search, discovery feeds, home feed, trending leaderboards, and creator leaderboards. You will inherit a production system serving live traffic and set the technical direction for this new area.
Key responsibilities include:
- Design and evolve the ranking stack, including engagement signal pipelines, decay and freshness models, and patterns combining pre-computed index-time signals with live re-ranking at query time.
- Make relevance a measured discipline by defining quality metrics per surface, building offline evaluation and golden-set tooling, and running online experiments to validate ranking changes.
- Own OpenSearch in production: index and mapping design, reindex and cutover safety, query performance, cost optimization, and operational headroom for discovery clusters.
- Build candidate generation and personalization for recommendation surfaces, partnering with data and ML teams on feature pipelines and models.
- Work directly with Product and Trust & Safety to define quality gates, cold-start behavior for new creators and characters, and safety constraints.
- Set engineering standards for the area: instrumentation, flag-gated rollouts, and runbooks enabling other teams to modify ranking behavior safely.
- Act as the technical point of contact for discovery across backend teams and mentor engineers whose work touches this area.
Requirements:
- 10+ years building production software with substantial depth in search, ranking, recommendations, or ML-driven relevance on consumer-scale products.
- Hands-on expertise with a production search engine (OpenSearch, Elasticsearch, Lucene, Vespa, or similar), covering index and analyzer design, relevance tuning, and cluster operations.
- Rigorous approach to relevance measurement: owned offline evaluation, golden sets, and A/B or interleaving frameworks; can explain what moved metrics and why.
- Fluency across the recommender stack (candidate generation, feature pipelines, embedding retrieval, re-ranking) with understanding of serving constraints that affect online performance.
- Strong distributed systems and backend fundamentals with production experience in Go or comparable systems language; comfort with data layers (SQL warehouses, streaming and batch pipelines).
- Demonstrated ownership of ambiguous areas: picked up something with no dedicated owner, decided what mattered, and delivered results.
- Clear written communication and judgment to work with Product on tradeoffs between relevance, freshness, safety, and latency.
- Preferred: experience with AI-generated or rapidly growing content catalogs, severe cold-start conditions, or ranking under trust-and-safety constraints.
