sqpe.devContribute
← Back to research
2026 / Research engine

Sema

Sema: A High-performance System for LLM-based Semantic Query Processing

What it contributes

Integrates semantic expressions into SQL and combines logical rewrites with adaptive reordering, fusion and prompt batching.

How to read the evidence

Preprint results cover 20 queries and selected baselines; no independent reproduction is claimed here.

We checked the primary publication record and abstract. This profile summarizes author-reported work; it does not claim a full-paper audit, implementation audit, or independent reproduction.

Explore the connections

SQLDuckDBAdaptive execution
  • FlockMTL — SQL / DuckDB extension
  • UQE — UQL / SQL dialect
  • SUQL — SQL extension