Staff/Senior Software Engineer, Hybrid IR and Systems

Helios Intelligence Platforms•New York City, NY

5d•Onsite

About The Position

Helios is building a new kind of company to solve America’s hardest problems, starting with the government interaction layer. Government shapes every consequential market, but the infrastructure connecting public institutions and private organizations remains fragmented, manual, and difficult to navigate. Helios is rebuilding that layer. Our core platform, Proxi, gives organizations the intelligence they need to understand what government is doing, why it matters, and what to do next. From that foundation, we design and deploy secure, mission-specific systems for government agencies, enterprises, and institutions operating in complex and highly regulated environments. We bring together frontier AI, deep public-sector expertise, and forward-deployed execution. Our team includes leaders and builders from the White House, U.S. Department of State, Datadog, and Microsoft. We are backed by leading institutional investors and trusted by organizations working on high-stakes problems across government and industry.

Requirements

Retrieval mechanisms and their complementarity.
Analyzers, tokenization, field boosts and BM25 parameterization for lexical indices.
HNSW parameters, quantization, shard design and recall against latency for vector indices.
Rank fusion, score normalization, learning to rank, cross encoder reranking, duplicate suppression and source coverage limits.
Query intent classification, entity linking, alias and acronym graphs, controlled expansion and translation of natural language into structured filters and facets.
Temporal modeling, valid vs. transaction time, supersession, corrections and effective date intervals.
Permission aware retrieval, including tenant isolation, PII filtering and cache key authorizations.
Relevance evaluation under sparse labels.
Multilingual and cross lingual retrieval.

Nice To Haves

Domain adaptation of embedding models through fine tuning, hard negative mining or distillation.
Work on policy, legislative, procurement, legal or healthcare corpora, where traceable information systems are mandatory.
Experience with transliteration and identifying entity names that should never be translated.
Operating search in classified, air gapped or forward deployed environments.
Open source contributions to search and ranking projects.

Responsibilities

Own Proxi’s retrieval stack end to end, from corpus onboarding and index design through hybrid retrieval, fusion, reranking, reindexing, and backfills.
Advance our query-understanding and temporal-retrieval capabilities, including intent detection, entity resolution, structured filtering, and jurisdiction-, date-, and version-aware search.
Improve retrieval performance, reliability, and cost through capacity planning, caching, reranker optimization, and safe degradation under load.
Expand federated search across Proxi’s separate data collections, including source-aware query planning, filtering, score calibration, and result merging.
Enhance time as a first-class retrieval concept so users can search the correct version, status, or historical state of a record.
Improve the dynamic connection system to map documents and records to the people, organizations, agencies, events, and relationships they concern.
Expand support for multimodal content searchable through the same retrieval interface as Proxi’s continuously ingested datasets.
Own relevance quality and production performance, including evaluation, indexing, observability, authorization, provenance, and recovery.