Extracts transactional document data template-free utilizing a proprietary transactional large language model.
Ingests documents across multiple ingestion routes including emails, shared drives, and PEPPOL networks.
Validates extracted line-item information against internal business rules and master data lists.
Automates communications for vendor exception handling using custom generative AI email drafts.
Why it works
Proprietary LLM delivers high out-of-the-box extraction accuracy without template setups.
Continuous human-in-the-loop validation updates learning weights natively with every human keystroke.
Supports multi- Subsidiary operations with granular permission schemes and localization features.
Strict enterprise compliance frameworks including SOC2, ISO 27001, and HIPAA protocols.
Rossum is an AI-first, cloud-native intelligent document processing (IDP) platform designed to automate transactional paperwork. Powered by its proprietary transactional LLM, Rossum Aurora, the platform ingests unstructured documents like invoices, purchase orders, and packing lists across multiple channels. It extracts and validates data template-free, handles complex multi-way matching, and integrates with major ERP systems to drastically reduce manual entry overhead.
Founded year:2017
Founder:Tomas Gogar, Petr Baudis, Tomas Tunys
Team size:100+
Popularity:Market leader in the enterprise intelligent document processing and AI-driven finance automation market
HQ:London, England, United Kingdom
Status:Acquired
Funding status:Acquired
Customer type:B2B
Funding:$104.6 million raised prior to acquisition
Pricing:Subscription
Tech stack:Python, React, TypeScript, Node.js, PostgreSQL, Amazon Web Services
Platform:Web
Integrations:SAP, Coupa, NetSuite, Workday, Microsoft Dynamics, Oracle, QuickBooks
What it does
Extracts transactional document data template-free utilizing a proprietary transactional large language model.
Ingests documents across multiple ingestion routes including emails, shared drives, and PEPPOL networks.
Validates extracted line-item information against internal business rules and master data lists.
Automates communications for vendor exception handling using custom generative AI email drafts.
Who it's for
Accounts Payable Departments
Supply Chain Operations Teams
Global Shared Service Centers
Logistics and Procurement Managers
Why it works
Proprietary LLM delivers high out-of-the-box extraction accuracy without template setups.
Continuous human-in-the-loop validation updates learning weights natively with every human keystroke.
Supports multi- Subsidiary operations with granular permission schemes and localization features.
Strict enterprise compliance frameworks including SOC2, ISO 27001, and HIPAA protocols.
Founder story
Founded in 2017 in Prague, Czech Republic by AI PhD students Tomas Gogar, Petr Baudis, and Tomas Tunys, Rossum targeted structural data limitations. Recognizing legacy optical character recognition (OCR) systems were fragile because they relied heavily on templates, they built a semantic understanding layout. The platform scaled into a leading global document pipeline until its acquisition by Coupa in 2026.
Alternatives
Comparison overview
Unlike legacy OCR platforms that require custom template maps for every single vendor layout, Rossum leverages its template-free transactional LLM to extract data seamlessly.
Compared to general document processing options, Rossum offers tailored accounts payable automation including 2/3/4-way matching capabilities natively out of the box.