Commercial agreements have always been the ultimate source of truth for businesses. Contracts and other written documents define the intricate financial, legal, and practical relationships between businesses. While many business documents have transitioned to digital form, the vital information in them is still extremely difficult to access, buried in a .pdf file or other unstructured digital format.
Thanks to advances in AI, cloud, imaging, NLP and other technologies, it is now finally possible to unlock all the business-critical information contained in your documents and use it to drive your business processes. The issue we hear relentlessly from customers has been the practical deployment of such promise, with no investment in data preparation, disruption of how people work with their favorite productivity tools, or impact on deployed line-of-business systems.
We have made it easy. Our document engineering approach has led us to design and deliver a product for frontline office workers and managers, and on small numbers of documents, without requiring lengthy training and preparation. Businesses of any size can tap directly into their contracts to dramatically improve their strategic insights, productivity, and compliance.
Digital technologies have made it easy to churn out large numbers of documents. However, the resulting text information and creative visual formatting continues to be nearly impenetrable and indecipherable to computers. The information must be manually extracted into spreadsheets, or other structured business systems to be useful. Many of the complex legal and business terms never make it into any structured form at all, so people must go back repeatedly to the written, digital documents on their computer.
This is not fun. It can be a ‘bespoke’ process, requiring a great deal of attention to the intricacies of highly varied, detailed documents. You may need one set of information for a specific purpose today, and something very different tomorrow, requiring another round of time-consuming and detailed work. It is expensive. Plowing through documents takes time away from everything else that makes the business run and grow.
Bottom line: the difficulty of getting critical information out of written documents represents a huge drag on business performance.
That’s why advances in document engineering have the potential to transform your business. Getting the right information in the right place at the right time can drive great strategic insights, timely decisions, crisp execution, and improved regulatory and business accountability.
Companies have been trying to solve this problem for decades, but the technology just was not advanced enough. Unstructured or semi-structured legal and business narratives are useful, even vital, for people to create and use, but are chaotic for computers. Information may appear to be structured, but the next document for the same purpose can structure similar information very differently. And the variable information often turns out to be the most important.
Text search, text extraction and data mining tools do allow you to find some things. Names, dates, dollar amounts, words, are easy to find – but difficult to interpret. With six monetary amounts in a contract, which one is the price, a monthly payment, a penalty, a cap on liabilities, a surcharge, a maintenance charge, the thing you are looking for? One contract calls the counter party a Provider, another calls them a Partner, another a Supplier. Are these similar relationships, or different? Contract Analytics and the various ‘Insight Engine’ or Enterprise Search offerings often miss these important distinctions.
There is both Art and Science involved as Attorneys craft a contract. What is an obligation? Something that has the word ‘Shall’ in front of it? Or a section that is much more sophisticated? Computers can handle the former, the latter not so much. As published for the American Bar Association, “An entire industry of attorneys, counsels, paralegals and consultants are busy trying to sort out the world, often accused of a dark art with complex terms and an impenetrable language of its own.”
Your company’s ‘small data’ can play a big role for your organization. Most organizations or divisions in large companies have dozens, or perhaps hundreds, of important existing contracts, not tens of thousands. Your contracts and other documents carry unique, proprietary terms and conditions. This ‘small data’ can be quite expensive and time-consuming to go through manually, repeatedly, to report out the details.
‘Big Data’ from the outside can introduce ‘Generic Data’ as well. When AI models are trained on outside datasets (Big Data) they will not be attuned to a company’s practical reality. Research is also finding that Big Data models can be ‘riddled with errors.’ Also, insights from one company’s data can be essentially ‘stolen’, by simply using them to feed the big data models for others. The Small Data approach respects the uniqueness and private nature of every company’s agreements.
To practically deploy AI to help, the AI model needs to be able to operate accurately on these relatively ‘small,’ complex, unstructured, unique, sets of documents.
Docugami applies our unique combination of the most sophisticated AI techniques to this set of problems: opaque files, natural language understanding, finding structure in chaos, associating information with the surrounding context, working with small datasets and more. The result is that every document is represented by a ‘shadow’ technology version, that computers can understand.
Once computers can understand the intricacies of text communications, they can help users in a host of new ways. For example, a user can simply open a single contract and click on any important terms, and immediately generate a report or export of that information for ALL similar contracts. Software can create dashboards for review: for renewal, for action, for compliance, for customers, for any other purpose of the moment. Information can flow through the organization: to assure performance, to check compliance or risk, to follow through on obligations, to calculate differences, to seek insights, to send out notifications, or insert into other business systems, Manual steps for all this can be dramatically reduced or eliminated.
Some organizations already tell us they can save 90% of the manual document work during their busiest renewal times, so they can focus on selling, and supporting their customers, to win and keep more business. The ROI is immediate.
The information from contracts is directly available to transform your business. By unlocking this information and making it accessible to computer systems, contracts become the business intelligence starter for your automation journey, instead of historical archives to sort through every day.