Synthetic enterprise dataset designed for benchmarking, evaluation, and testing of AI agents, RAG (Retrieval-Augmented Generation) systems, enterprise compliance auditing, and search engines.
This repository contains realistic synthetic corporate documents spanning various enterprise domains:
- CEO & Executive Communications: High-level emails, strategic directives, and acquisition term sheets.
- Engineering & Operations: Technical documentation, system specs, and operational logs.
- Finance & M&A: M&A term sheets, financial reports, and transactional records.
- HR & HR Records: Employee records, policy documents, and administrative forms.
- Legal & Compliance: Contracts, vendor SLAs, legal settlement documents, and regulatory compliance files.
- Medical & Patient Archives: Synthetic clinical records, patient archives, and medical R&D data.
- Research & Logistics: R&D reports, logistics notes, and supply chain records.
.
├── CEO_Email_BiotechX_Titan.txt
├── Engineering/
├── Finance/
├── Finance_and_M_A/
├── HR/
├── HR_Records/
├── Legal/
├── Legal_Case_384_Victim_Settlement.pdf
├── Legal_Compliance/
├── Legal_Contracts/
├── Medical/
├── Patient_Archives/
├── Project_Titan_Acquisition_TermSheet_v7_Draft.docx
└── Research_and_Logistics/
This dataset is ideal for:
- Benchmarking enterprise RAG pipelines (e.g., hybrid vector/keyword search, graph-RAG).
- Evaluating LLM document understanding across standard formats (
.txt,.pdf,.docx). - Testing AI compliance enforcement tools and data governance frameworks.
This dataset is composed entirely of synthetic / mock data generated for testing, evaluation, and demonstration purposes.