Query & Algorithms Engineer

About the role

Purple8's query layer combines a cost-based planner with graph-aware vector retrieval so that hybrid search stays both accurate and efficient at scale. You'll own the planner, the graph algorithm suite, and the vector integration that make retrieval work.

What you’ll own

• Own the cost-based query planner and its strategy selection.
• Maintain and extend the graph algorithm suite (PageRank, communities, paths, centrality, link prediction).
• Advance graph-aware vector retrieval and its efficiency.
• Push filters into the planner rather than post-filtering in application code.

What we’re looking for

• Strong algorithms and data-structures fundamentals.
• Experience with query planning or graph processing at scale.
• Python performance engineering; comfort operating directly on a storage iterator, not in-memory lists.

Nice to have

• ANN / HNSW internals.
• Sparse linear algebra (scipy) for spectral methods.