Agentic Search Infrastructure Engineer - Moveworks

Job description

Contribute to the org-wide shift from product-specific search solutions to a composable, agent-native infrastructure foundation that agents and applications build on to locate, search, discover, and recommend resources across heterogeneous enterprise data. Build and own the infrastructural substrate beneath the agentic search platform, spanning the index engine, the ingestion and enrichment pipeline, and the entity, identity, and feature intelligence that gives content context.

Solve the hard problem of keeping index-time enrichments aligned with what agents expect at query time, so retrieval, ranking, and reasoning all build on a consistent foundation. Build and operate hybrid indexing technology at scale across large clusters, spanning keyword-based sparse-vector indices, embedding-based semantic indices, and emerging semantic-ID indexing driven by next-token prediction. Own the full index lifecycle, and expose clean, unified contracts to the search primitives and recipes that agents compose on top of the corpus.

Design and run high-throughput ingestion and enrichment pipelines that produce a multi-representation, permission-aware corpus at index time. Orchestrate the stages that make content agent-ready, including parsing, chunking, embedding, entity resolution, and access control, and evolve them as new agentic use cases emerge. Build and operate multi-modal indexing across the full mix of enterprise data, from free text and structured content such as tables to diagrams, images, and audio and video.

Represent each modality so it is retrievable through a unified surface, letting agents find the right material regardless of its form or source. Contribute to a knowledge graph that resolves entities across enterprise data sources, letting agents link disparate resources such as tickets, files, and people through canonical entity relationships. Improve the performance, scalability, and observability of search, including query latency, indexing throughput, and cluster health, to deliver low-latency, high-quality agentic search experiences.

Apply techniques such as vector quantization to balance model quality, retrieval performance, and storage cost. Own search infrastructure end to end, from intake and design through rollout and ongoing operations, and help shape the search roadmap as the platform grows to support new agents, primitives, and recipes, including needs not yet anticipated. Drive reliability and operability across the platform through index lifecycle policies, capacity planning, and cluster scaling strategies that keep a growing library of reusable search capabilities healthy in production.

This is a highly collaborative, in-person role, working cross-functionally with Search Ranking, ML, and Platform engineering teams to deliver robust, production-grade agentic search capabilities.

Qualifications

BS or MS in Computer Science or an equivalent engineering background; PhD is a plus. 1+ years of full-time industry experience. Hands-on experience with search technologies; proficiency in one or more search technologies such as Lucene, Solr, Elasticsearch, or OpenSearch is required, with additional experience in vector databases being an advantage. Excited to be part of an in-person team culture, collaborating on-site with colleagues every day.