OfficeQA Pro: An Enterprise Benchmark for End-to-End Grounded Reasoning
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2026
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866911499287527424 |
|---|---|
| author | Opsahl-Ong, Krista Singhvi, Arnav Collins, Jasmine Zhou, Ivan Wang, Cindy Baheti, Ashutosh Oertell, Owen Portes, Jacob Havens, Sam Elsen, Erich Bendersky, Michael Zaharia, Matei Chen, Xing |
| author_facet | Opsahl-Ong, Krista Singhvi, Arnav Collins, Jasmine Zhou, Ivan Wang, Cindy Baheti, Ashutosh Oertell, Owen Portes, Jacob Havens, Sam Elsen, Erich Bendersky, Michael Zaharia, Matei Chen, Xing |
| contents | We introduce OfficeQA Pro, a benchmark for evaluating AI agents on grounded, multi-document reasoning over a large and heterogeneous document corpus. The corpus consists of U.S. Treasury Bulletins spanning nearly 100 years, comprising 89,000 pages and over 26 million numerical values. OfficeQA Pro consists of 133 questions that require precise document parsing, retrieval, and analytical reasoning across both unstructured text and tabular data. Frontier LLMs including Claude Opus 4.6, GPT-5.4, and Gemini 3.1 Pro Preview achieve less than 5% accuracy on OfficeQA Pro when relying on parametric knowledge, and less than 12% with additional access to the web. When provided directly with the document corpus, frontier agents still struggle on over half of questions, scoring 34.1% on average. We find that providing agents with a structured document representation produced by Databricks' ai_parse_document yields a 16.1% average relative performance gain across agents. We conduct additional ablations to study the effects of model selection, table representation, retrieval strategy, and test-time scaling on performance. Despite these improvements, significant headroom remains before agents can be considered reliable at enterprise-grade grounded reasoning. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_08655 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | OfficeQA Pro: An Enterprise Benchmark for End-to-End Grounded Reasoning Opsahl-Ong, Krista Singhvi, Arnav Collins, Jasmine Zhou, Ivan Wang, Cindy Baheti, Ashutosh Oertell, Owen Portes, Jacob Havens, Sam Elsen, Erich Bendersky, Michael Zaharia, Matei Chen, Xing Artificial Intelligence Computation and Language Information Retrieval We introduce OfficeQA Pro, a benchmark for evaluating AI agents on grounded, multi-document reasoning over a large and heterogeneous document corpus. The corpus consists of U.S. Treasury Bulletins spanning nearly 100 years, comprising 89,000 pages and over 26 million numerical values. OfficeQA Pro consists of 133 questions that require precise document parsing, retrieval, and analytical reasoning across both unstructured text and tabular data. Frontier LLMs including Claude Opus 4.6, GPT-5.4, and Gemini 3.1 Pro Preview achieve less than 5% accuracy on OfficeQA Pro when relying on parametric knowledge, and less than 12% with additional access to the web. When provided directly with the document corpus, frontier agents still struggle on over half of questions, scoring 34.1% on average. We find that providing agents with a structured document representation produced by Databricks' ai_parse_document yields a 16.1% average relative performance gain across agents. We conduct additional ablations to study the effects of model selection, table representation, retrieval strategy, and test-time scaling on performance. Despite these improvements, significant headroom remains before agents can be considered reliable at enterprise-grade grounded reasoning. |
| title | OfficeQA Pro: An Enterprise Benchmark for End-to-End Grounded Reasoning |
| topic | Artificial Intelligence Computation and Language Information Retrieval |
| url | https://arxiv.org/abs/2603.08655 |