Senior AI/ML Software Engineer

Recruiting From Scratch•Boston, NY

3d•Hybrid

About The Position

The Senior AI/ML Software Engineer will join the Discovery Pod, which builds the ranking, recommendation, and search systems used to find diamonds and gemstones across the company's marketplace and individually branded jeweller storefronts. The role involves production information retrieval and applied machine learning, including search infrastructure, embedding-based retrieval, learning-to-rank, LLM-powered contextual and conversational search, and image-embedding-based search. The engineer will own models end to end, from features and training through evaluation, deployment, and monitoring, and will design systems with an understanding of how probabilistic models behave and fail. The role is described as hybrid in Boston and New York, with remote work available on the East Coast. Boston does not yet have an office, though a hub is planned; New York has an office in Midtown. The company has a global remote team and no core working hours. The position is on a small, high-ownership team building AI-powered products. Compensation is $220,000–$300,000 base, depending on seniority, location, and performance, plus equity. The technology environment includes OpenSearch, PostgreSQL, Spark/EMR/Glue ingestion jobs, Snowflake and dbt, TypeScript/Node.js, GraphQL, Nx, PNPM, Keycloak, Okta, Jenkins, GitOps, AWS, Kubernetes-based deployments, and Grafana. Embedding-based retrieval, ranking models, and image-embedding search are new builds for the Discovery role.

Requirements

5–10 years of experience.
Experience building or operating information retrieval infrastructure, such as Elasticsearch/OpenSearch, Vespa, or vector-search databases such as pgvector, Pinecone, or Weaviate.
End-to-end applied machine learning experience in ranking and recommendations, including learning-to-rank, embedding-based retrieval, and collaborative filtering, with ownership of features, training, evaluation, deployment, and monitoring.
Production experience building applications on language models, including prompt design, retrieval-augmented approaches, and graceful handling of zero-result and tradeoff scenarios.
Experience building and optimizing low-latency, real-time model-serving systems under production load.
Specialized search- and ranking-quality evaluation experience, including offline ranking metrics and interleaving experiments.
Hands-on experience building with foundation models and LLMs, including evaluating and stress-testing their output.
Experience designing graceful failure handling for probabilistic systems and accounting for cases where models are confidently wrong.
Comfort using AI coding agents to build ML infrastructure and pipelines, with strong evaluation and testing discipline.

Nice To Haves

Computer vision fundamentals, especially image embeddings, and experience building image-to-attribute or image-to-product-match pipelines.
Experience designing multi-tenant, configurable systems that behave differently per tenant without forking the codebase.
Experience building conversational or assistant-style search products.
Background on a marketplace or e-commerce company's search, ranking, or recommendations team.

Responsibilities

Build and operate search infrastructure to power ranking and multimodal search across the marketplace.
Design multi-tenant, configurable ranking and recommendation systems for the marketplace and branded jeweller storefronts without forking the codebase.
Build and improve contextual search and a planned conversational assistant using prompt design and retrieval-augmented approaches, handling zero-result and tradeoff scenarios gracefully.
Build image-to-attribute and image-to-product-match pipelines using image embeddings for image-upload search.
Design recovery paths and confirmation steps for probabilistic systems, especially before high-stakes actions such as generating a manufacturing quote from an interpreted image.
Define and run search- and ranking-specific evaluations, including offline ranking metrics and interleaving experiments.
Evaluate and stress-test foundation-model and LLM output as part of the engineering process.
Own the full lifecycle of ranking and recommendation models, including features, training, evaluation, deployment, and monitoring.
Design and optimize low-latency, real-time model-serving systems for production workloads.
Use AI coding agents to build ML infrastructure and pipelines, supported by rigorous evaluation harnesses that identify ranking and retrieval bugs.