Senior Search Engineer, Index Team - praca zdalna

Firma Kagi Inc szuka osoby na stanowisko Senior Search Engineer, Index Team: poziom senior, czyli decyzje podejmuje się tu samodzielnie. Oferta należy do kategorii „IT i programowanie” i jest w pełni zdalna. Firma nie ogranicza miejsca zamieszkania kandydata, więc zgłoszenie można wysłać z dowolnego miejsca.

Firma nie podała żadnej kwoty, ustala się to na rozmowie. Ogłoszenie nie stawia żadnych warunków co do godzin pracy.

Ta oferta przeszła automatyczną kontrolę: na listę nie trafiają ogłoszenia wymagające zagranicznego pozwolenia na pracę, sponsorowania wizy, określonego obywatelstwa ani zamieszkania we wskazanym kraju.

Senior Index Team Engineer About Kagi Kagi is your companion for a better web. We're building a friendly internet where you can explore, learn, and interact with confidence - free from ads, tracking, and distractions. Kagi builds advanced, user-centric technology to empower you, not replace you. Our products include Kagi Search, Orion browser, Kagi Assistant, Kagi News, Kagi Summarizer, and Kagi Translate. Every product reflects our commitment to a surveillance-free internet, designed around the user.

About the Index Team The Index team builds and maintains Kagi's own search index - the layer that makes our results meaningfully different from everyone else's. While most search engines stop at aggregating results from the same sources, we go further: we build, maintain, and continuously improve our own index to give Kagi users results that nobody else can offer. It's a hard technical problem involving distributed job scheduling, large-scale data pipelines, index quality engineering, and real-time monitoring - and the work directly shapes what millions of searches return.

The Role We're looking for a Senior Engineer to work on the systems that power Kagi's own index. You'll own pieces of the full pipeline - from URL frontier and job scheduling through document ingestion, processing, and indexed retrieval - and help us make that pipeline faster, fresher, and more reliable. You'll also work on index quality: understanding what's in the index, what should be, and what shouldn't be. This is a role for someone who finds deep satisfaction in the unglamorous work that makes a search index actually good: the scheduling logic that's wrong 0.1% of the time, the deduplication that keeps coverage clean, the monitoring that catches quality regressions before users do.

What You'll Do

* Embed within the Search team to ensure index improvements translate into better results for user.

* Design and improve the URL frontier and crawl job scheduling system - how we decide what to index, when, and in what order

* Own and evolve the indexing strategy (what we index post-crawl) and ranking implementation.

* Investigate and fix index quality issues - spam, stale content, deduplication, coverage gaps

Must-Haves

* 5+ years of backend engineering experience in production information retrieval/search systems

* Prior experience with search index infrastructure, URL frontier systems, or crawl scheduling

* Deep expertise across the information retrieval stack: index design, relevance scoring and ranking optimization, near-duplicate detection, query engine architectures, and graph-based signal extraction.

* Strong database skills - you're comfortable designing and optimizing complex PostgreSQL schemas and queries, including large job/task tables that need to scale

* Proficiency in Python and/or Rust - our primary languages for this work

* Ability to reason about large-scale data pipelines: throughput, correctness, idempotency, and operational reliability

Nice to have:

* Experience with observability tooling (Prometheus, Grafana or similar)

* Comfort with Redis for distributed state and queue management

* Experience with distributed job scheduling or work queue systems - you understand the failure modes and edge cases

* Experience operating systems where correctness and data quality are the primary constraints

If you've ever spent a week tracking down a subtle bug in a distributed job scheduler and felt proud rather than defeated at the end of it, this is the role. Originally posted on Himalayas