Search Engineer

About this role

Engineering builds and runs TalentOS — the hiring platform this careers site is itself running on. Job search, the application flow, the recruiter console, and the services underneath all of it.

Own job search: indexing, query understanding, ranking and the location logic that decides whether someone in Virar sees the role in Andheri. Relevance here is somebody finding work.

The platform is multi-tenant and the correctness bar is unusually high for a product of its size: an application that goes missing is somebody's career, not a support ticket. Traffic is spiky by nature — a campus placement season concentrates a year of applications into a fortnight — so capacity and queueing are ordinary design concerns rather than scaling exotica.

Primary responsibilities

• Own job search end to end — the index, the query pipeline, the ranking and the relevance
• Build the analysis chain that makes a search work in practice: synonyms across job-title vocabulary, stemming, and the transliteration that matters when people type Indian place names several different ways
• Handle geographic search properly, so a search around Mumbai reaches a posting in Virar that never says "Mumbai"
• Design and run relevance evaluation — a judged set, offline metrics, and online tests that measure whether people found what they wanted
• Keep query latency inside its budget as the corpus and the query volume grow
• Build the facet and filter behaviour, including counts that stay correct when filters interact

Skills required

Day to day this role works with Python, Elasticsearch, Information Retrieval and Ranking.

• Four or more years on search or recommendation systems
• You measure relevance rather than arguing about it