Staff Software Engineer, Search Systems - Code Data

Mercor•San Francisco, CA

16h•$250,000 - $500,000•Onsite

About The Position

Mercor is a leading AI data company focused on organizing human intelligence to power the AI economy. They are building the infrastructure that connects human expertise with frontier AI models. Millions of domain experts use the platform daily, earning significant income to train advanced AI models. Mercor's APEX benchmark family assesses the real-world impact of AI on professional tasks, and Mercor Enterprise extends this capability to Fortune 500 companies, enabling them to capture and translate their employees' expertise into AI agents. The company operates in a fast-paced, ambitious environment, working with researchers, operators, and AI companies at the forefront of AI development. Mercor is a profitable Series C company with a $10 billion valuation and requires in-person work five days a week in their San Francisco, NYC, or London offices.

Requirements

8+ years of professional software engineering experience, including 3+ years operating at a Senior level or above, with a Staff-level track record of org-wide technical impact.
Deep, hands-on expertise building search and retrieval systems: dense-embedding retrieval, lexical scoring (BM25/TF-IDF), hybrid ranking, and re-ranking.
Strong understanding of the search algorithms and index internals—vector/ANN indices (e.g. HNSW, IVF, product quantization), inverted indices, and engines such as Elasticsearch/OpenSearch, Lucene, FAISS, or vector databases.
A track record of making the right cost-vs-speed tradeoffs: latency budgets, throughput, memory footprint, and infrastructure spend on high-QPS systems.
Familiarity translating natural-language questions into structured search queries (query understanding, semantic parsing, or LLM-assisted query generation).
Excellent systems fundamentals: distributed systems, data modeling, and API design at scale.
Demonstrated technical leadership and mentorship—you've helped junior and senior engineers grow and level up an engineering team.
Genuine excitement for agentic development and new technology, fluency with modern AI dev tools (e.g. Claude Code, Cursor, Copilot), and a deep passion for writing great code.
Excellent communication—able to make complex tradeoffs legible to both engineers and leadership. Strong opinions, loosely held. High ownership, pragmatism, and a bias toward shipping.

Nice To Haves

Experience with code search or code understanding—retrieval over code, code embeddings, or working with code-specific models—and an appreciation for why matching similar code tasks is harder than matching text.
Experience training or fine-tuning code embedding models or code-specific LLMs.
Experience with learning-to-rank, semantic search, or recommendation systems in production.
Familiarity with LLM-based retrieval, RAG patterns, and model routing/selection.
Background operating latency-critical services on modern cloud and orchestration infrastructure.

Responsibilities

Own the architecture of Mercor's code search and retrieval systems end to end—hybrid retrieval combining dense code embeddings and BM25, candidate generation, ranking, and re-ranking over code tasks.
Solve the hard problem of identifying similar code tasks—retrieval that captures code structure, semantics, intent, and difficulty rather than surface text.
Build the system that identifies and selects the correct code-specific models for a given task, and routes tasks to the right model.
Design natural-language-to-query translation that turns NLP questions into precise search over code and tasks.
Design and operate the indexing pipeline so the task index stays fresh and consistent as new tasks, solutions, and results arrive continuously—balancing incremental updates, full rebuilds, and real-time ingestion.
Make the cost-and-speed tradeoffs that keep search fast and economical at scale: embedding dimensionality and quantization, ANN index choice and parameters, caching, sharding, and serving infrastructure.
Build the systems and evaluation harnesses that let us continuously evolve embeddings, models, and search quality—safely swapping in new code models, re-embedding corpora, and A/B testing relevance as SOTA advances.
Define and drive the long-term technical strategy for code retrieval across the organization, and lead the highest-stakes design reviews.
Establish evaluation metrics, offline/online testing, and quality guardrails so search improvements are measurable and regressions are caught before they ship.
Stay deeply hands-on: prototype critical systems, ship production code, and unblock teams on their hardest retrieval and infrastructure problems.
Mentor and grow engineers—junior and senior—through design reviews, pairing, and clear technical writing, raising the technical bar across the org.
Partner with product, researchers, and engineering leadership on build-vs-buy decisions, platform investments, and technical hiring.