Rossum Named a Strong Performer in The Forrester Wave™: Document Mining And Analytics Platforms, Q2 2024

Such a proud moment. Rossum named a Strong Performer in The Forrester Wave™: Document Mining And Analytics Platforms, Q2 2024 report. Our first-ever inclusion. We believe recognition from Forrester acknowledges our expertise and commitment to achieving our dream of one person effortlessly handling one million transactions a year.

* This post was originally published 30/05/2024. The content has been updated with regard to data mining security and compliance, implementation, and what the future holds.

What is a Strong Performer?

Well… Rossum.

We rocked in with the highest scores possible in 2 criteria. Innovation and Gen AI – post-processing.

We believe being named as a Strong Performer by a leading research firm demonstrates that we’ve earned our place in the document mining and analytics market, following our mission of anticipating customer needs and maintaining a customer-centric strategy.

Less than 100% customer satisfaction is not an option.

The Forrester Wave™ report reads, “Rossum’s differentiated approach to leverage genAI in document mining, using a specialized fine-tuned model to validate both the results and confidence scores of the foundational LLM, clearly stands out among the competition.“

Forrester definition of document mining and analytics

“Software that applies document and text mining and analytics technology to extract information from semi-structured documents that may include forms and/or document sections.” Source: The Document Mining And Analytics Landscape Platforms, Q1 2024

Also known as document analytics, document understanding, or intelligent document extraction and processing, document mining and analytics uses natural language processing (NLP) built on machine learning-based AI to enable computers to understand, interpret, and generate human language.

Extracting data and text from documents and images such as names, places, dates, and numbers. Then using the information extracted to populate business documents and identify strategic insights.

Why read the Forrester Wave™ report

The Forrester Wave™ report helps businesses make informed decisions when choosing a technology solution. Saving months of research and stress, and helping enterprise companies understand which vendor will be the best fit for their specific use cases.

Why is document mining and analytics important?

Imagine if you could extract real-time insights from your transactional documents. Insights that would drive your business. Accelerate productivity by automating your document processing end to end.

Document mining and analytics is a process that evaluates and analyzes unstructured documents. Using natural language processing, the technology sorts and identifies patterns, topics, and keywords in data sets.

While document mining and analytics can be used on all types of business documents, we specialize in transactional documents such as invoices, purchase orders, and bills of lading.

This makes us unique.

This focused approach on transactional document automation differentiates us from generic document mining solutions.

Hey. It makes us leaders.

To achieve our mission, we built our advanced AI engine, Rossum Aurora. Powered by our proprietary Transactional Large Language Model. Trained on millions of annotated transactional documents, our LLM isn’t hampered by hallucinations or data leakage. And, it achieves human-level accuracy almost instantly, while providing enterprise-grade safety. According to the Forrester Wave™report our next-Gen AI is already delivering tangible value to our customers. 

“Reference customers praised the platform’s overall business value, the overall relationship with the vendor, and business user and data/analytics professional UX.” Source: The Forrester Wave™: Document Mining And Analytics Platforms, Q2 2024

Targeting transactional documents means we’re uniquely positioned with key differentiators…

AI-first and cloud-native

• Our AI-first platform ensures we deliver maximum scalability, agility, and speed of development 
• Cloud-native, we tap into vast datasets, constantly training our AI to become smarter over time 
• Our proprietary LLM ensures fast learning, unparalleled accuracy, and speed in automation

End-to-end automation

• Rossum does more than just document understanding. Sitting between transacting parties, we orchestrate the transactional process end to end, from reading and answering emails to document approvals

We work for with you

• Like a new team member, our software learns how you deal with your transactional documents, and figures out how to do it with you, offering a seamless Human-AI collaboration

Early users of Rossum Aurora such as allnex, NFI Industries, Port of Rotterdam Authority, and Wolt already seeing an increase in accuracy and productivity…

• 10x reduction in the number of documents needed to reach desired accuracy
• An average 37.6% error reduction, without those pesky hallucinations
• Accelerated discrepancy handling through generative AI custom emails straight into existing workflows

Rossum Aurora requires 10x fewer training examples to reach desired accuracy levels,
compared to previous AI generation.

“Rossum Aurora achieved a 92.6% accuracy rate after processing only 20 documents. To see the AI learning and instantly getting smarter gives us confidence we can truly deploy AI to automate processes and scale sustainably,” said Marco Favaretti, Financial Controller at Adyen, the global financial technology platform and one of Rossum Aurora’s early adopter customers.

Document mining implementation

Implementation starts with comprehensive infrastructure planning. Determine your document volume, types, and processing requirements to choose the most appropriate deployment model. 

Cloud deployments offer scalability and reduced maintenance, while hybrid solutions satisfy strict data residency requirements. 

Training

A robust training strategy should kick off with a representative sample of 50-100 documents per document type. Prioritize document types based on volume and business impact. Leading platforms now use active learning techniques, where the system identifies challenging cases for human review, reducing training time.

Change management 

The most crucial part of implementation is change management. You’ll need to adopt a phased approach, starting with a pilot program in one department or document type. This allows for…

• Testing and refinement of workflows
• Increasing user confidence through early wins
• Identifying and addressing process bottlenecks
• Creating internal champions and best practices 

Security and compliance in document mining

Document mining solutions must implement multiple layers of security to protect sensitive business data. At the infrastructure level, this means employing end-to-end encryption, using industry-standard protocols like TLS 1.3 and AES-256 encryption. 

For enterprises handling sensitive documents, role-based access control – RBAC – is crucial. Allowing you to define precisely who can access which documents and what actions they’re allowed to perform.

Compliance certification

Enterprise document mining solutions must maintain key certifications that demonstrate their commitment to security. These typically include…

• SOC 2 Type II compliance, which verifies the system’s security, availability, and confidentiality controls
• For global operations, ISO 27001 certification ensures adherence to international security standards
• Healthcare organizations processing medical documents require HIPAA compliance
• Financial institutions need solutions that comply with regulations like FINRA and SEC requirements

Data privacy measures

Modern document mining platforms incorporate privacy-preserving features such as automatic PII detection and redaction. Advanced systems use AI to identify sensitive information across multiple languages and document types, automatically masking or encrypting this data based on predefined rules. 

Your business should adopt granular control over data retention periods and be able to implement automated deletion policies to comply with regulations like GDPR’s right to be forgotten.

The future of document mining

The integration of specialized Large Language Models (LLMs) has revolutionized document mining capabilities. Unlike generic LLMs, domain-specific models trained on business documents achieve higher accuracy with fewer training examples. Models like our Transactional LLM. 

These models can understand complex document layouts, handle multiple languages, and adapt to new document types with minimal human intervention. 

The future of document mining lies in end-to-end process automation. Advanced systems have moved beyond simple data extraction to intelligent process orchestration. This includes automatic email response generation, smart routing of documents based on content, and predictive analytics that can forecast processing bottlenecks.

What does this mean for Rossum?

Being recognized as a Strong Performer in document mining and analytics strengthens our commitment to customer-centric innovation and development. To meet customer needs and exceed them.

To delivering unmatched productivity and accuracy for enterprise businesses. Eliminating manual data entry and transforming how businesses tackle the document chaos challenges they face. Ultimately, delivering a collaboration between AI and humans that accelerates performance, creativity, and innovation. 

This triumph is thanks to our customers, partners, and the entire Rossum team of experts. Huge thanks to each of you.