Train models to translate land use planning controls with aligned texts and domain-specific terminology.

Train models to assess chromatography compatibility by reasoning across method parameters, sample conditions, and instrument specifications.

Train multi-step reasoning about self-consumption, grid exports, and curtailment under commercial solar operating constraints.

Train models to generate energy optimization code from flow battery scheduling requirements and operating constraints.

Trace the code path from an incoming notification request through dispatch and status updates.

Build more traceable shipping document QA with source-grounded answers about dry bulk charterparties, laytime, and demurrage.

Diagnose possible triggers of new convection from satellite time-series clues, with evidence and uncertainty clearly identified.

Train models to calculate transport duration and remaining time and assess overdue deliveries using refrigerated transport event sequences with explicit supervised answers.

This dataset focuses on enhancing the safety of the elderly using stairs at home by providing support through intelligent Q&A.

Train models to assess commercial and industrial rooftop solar feasibility by reasoning across roof, load, structural, shading, and grid constraints.

Combine diverse rescue records with failure analyses to help models produce evidence-based, professional incident reviews.

Train models to assess factors affecting giant freshwater prawn growth and feed efficiency using farming context and evidence.

Evaluate whether a system can find the right service, identify the right merchant, and support its answer with evidence from multi-merchant catalogs.

Train models to assess offshore oil well intervention applicability and explain their reasoning from well facts and operational rules.

Ground dry bulk port and terminal operations answers in source webpages and evidence passages for more traceable RAG systems.

A high-quality text conversation dataset containing hotel ratings, reviews, and replies.

The key information extraction dataset for doctor-patient conversations is used to enhance the information extraction capability of medical dialogue systems.

Train models to derive indoor DAS topologies and explain coverage design decisions through structured problems and constraint analysis.

Train models to reason about feeding changes, growth outcomes, and confounding factors in trout farming scenarios.

Train models to understand traditional music practices and their cultural contexts with specialist text and verifiable evidence.