Enterprise RAG Internal Knowledge Search Benchmark Sample

A synthetic internal company workspace and evidence-linked benchmark table for evaluating enterprise retrieval-augmented generation systems. The dataset is authored by Karmane and was last updated on June 8, 2026. It is designed around a realistic fictional company, AsteraOps Cloud, with multiple departments and projects.

Use Cases

• Benchmarking retrieval-augmented generation systems based on the synthetic internal company workspace.
• Evaluating internal search assistants based on the evidence-linked benchmark table.
• Testing knowledge-base copilots using the realistic fictional company structure.
• Developing deep-search evaluation pipelines based on the packaged benchmark data.

Strengths

• Designed around a realistic fictional company, AsteraOps Cloud, with multiple departments and renamed projects.
• Packages both a synthetic workspace and an evidence-linked benchmark table into one product.

Limitations

• Description metadata is limited; actual data quality requires manual inspection after download.
• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.

Provenance

Source

huggingface

Collection Method

Synthetic generation, likely for benchmarking purposes.

Freshness

Last updated 2026-06-08 21:12:58; freshness should be verified.