{"ID":5675994,"CreatedAt":"2026-07-03T01:40:09.565152011Z","UpdatedAt":"2026-07-04T20:51:25.697068714Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2607.01468","arxiv_id":"2607.01468","title":"CADENZA in Action: Breaking the Monolith with Intent-Dependent Plan Spaces for Semantic Queries","abstract":"Semantic query processing engines execute semantic operators, whose behavior is specified by natural-language intents, via model inference over multimodal data. Most existing optimizers optimize the operators at the granularity of monolithic implementations -- such as LLMs and embedding models -- forcing a trade-off between expensive model calls and cheaper alternatives that fail to capture intent-dependent semantics. We present CADENZA, a semantic operator optimizer that compiles an intent into decomposed steps, selects concrete physical implementations for each step, and tunes their parameters under user-specified quality-latency-cost preferences. In this demonstration, users interact with CADENZA through a web interface over multimodal databases, exploring how an intent is decomposed into alternative plans, how each plan is optimized, and how different preferences yield different winning plans.","short_abstract":"Semantic query processing engines execute semantic operators, whose behavior is specified by natural-language intents, via model inference over multimodal data. Most existing optimizers optimize the operators at the granularity of monolithic implementations -- such as LLMs and embedding models -- forcing a trade-off be...","url_abs":"https://arxiv.org/abs/2607.01468","url_pdf":"https://arxiv.org/pdf/2607.01468v1","authors":"[\"Jaehyun Ha\",\"Yongjoo Park\",\"Wook-Shin Han\"]","published":"2026-07-01T21:00:40Z","proceeding":"cs.DB","tasks":"[\"cs.DB\"]","methods":"[\"Large Language Model\"]","has_code":false}