Software Engineer, Search Systems - Code Data
San Francisco • FullTime
Posted 1mo ago
Remote Work Policy
On-site
Employment Type
FullTime
Categories
Applied AI Engineer
About the job
Mercor is a leading AI data company organizing human intelligence to power the AI economy. We are building the layer between human expertise and frontier models, with millions of domain experts on our platform training AI models. This role focuses on code data, specifically owning the architecture and algorithms for searching across code tasks. The goal is to find similar tasks, route them to the right models, and translate natural language questions into precise retrieval queries. This is a hands-on technical leadership position where you will design and build retrieval systems, manage trade-offs for speed and cost at scale, and continuously evolve these systems as state-of-the-art code models advance. You will also set technical direction, mentor engineers, and raise the bar for engineering practices within the organization.
Responsibilities
- Own the end-to-end architecture of Mercor's code search and retrieval systems, including hybrid retrieval, candidate generation, ranking, and re-ranking.
- Solve the challenge of identifying similar code tasks by developing retrieval that understands code structure, semantics, intent, and difficulty.
- Build systems to identify and select appropriate code-specific models for tasks and route tasks accordingly.
- Design natural-language-to-query translation to convert NLP questions into precise search queries.
- Design and operate the indexing pipeline to ensure the task index remains fresh and consistent with continuous updates.
- Make cost-and-speed trade-offs for search systems at scale, considering embedding dimensionality, ANN index choice, caching, sharding, and serving infrastructure.
- Build systems and evaluation harnesses for continuously evolving embeddings, models, and search quality, including safe swapping of new models and A/B testing.
- Define and drive the long-term technical strategy for code retrieval across the organization.
- Establish evaluation metrics, testing methodologies, and quality guardrails for search improvements.
- Remain hands-on by prototyping critical systems, shipping production code, and unblocking teams on retrieval and infrastructure challenges.
- Mentor and grow engineers through design reviews, pairing, and technical writing.
- Partner with product, researchers, and engineering leadership on build-vs-buy decisions, platform investments, and technical hiring.
Requirements
- 8+ years of professional software engineering experience, with at least 3 years at a Senior level or above.
- Staff-level track record of organization-wide technical impact.
- Deep, hands-on expertise in building search and retrieval systems, including dense-embedding retrieval, lexical scoring (BM25/TF-IDF), and hybrid ranking/re-ranking.
- Strong understanding of search algorithms and index internals, including vector/ANN indices (e.g., HNSW, IVF, product quantization), inverted indices, and engines like Elasticsearch/OpenSearch, Lucene, FAISS, or vector databases.
- Proven track record of making cost-vs-speed trade-offs for high-QPS systems (latency, throughput, memory, infrastructure spend).
- Familiarity with translating natural-language questions into structured search queries (query understanding, semantic parsing, or LLM-assisted query generation).
- Excellent systems fundamentals, including distributed systems, data modeling, and API design at scale.
- Demonstrated technical leadership and mentorship experience.
- Experience with code search or code understanding (retrieval over code, code embeddings, code-specific models) is a plus.
- Appreciation for the complexities of matching similar code tasks compared to matching text is a plus.