AI has arrived. Long live AI. Now what do I do with it?

AI has arrived. Long live AI. Now what do I do with it?

Over the last year many of us have consumed AI news at firehose speeds. Every week there is a new model, a new framework, a new best practice and a new guru. One of the biggest challenges companies are dealing with is the why of it all. Why should I care about all this innovation? How do I apply it in a way that will meaningfully impact my business, my employees and my customers. In this article we’ll explore some of the early use cases for Nomad Data’s Document Search that we’re seeing out in the wild.

Nomad Data’s Document Search is an enterprise document processing engine, which allows users to perform a variety of tasks against large document repositories involving information lookup, summarization and extraction. When we first launched Document Search, we had a few use cases in mind but have seen customers continually coming up with creative new ways to take advantage of it. Below are some of the most creative use cases we’ve seen so far.

Vendor Contract Data Extraction

Buried in vendor agreements is a wealth of information including purchase dates, vendor names and addresses, product purchase details, renewal information and much more. Our clients are often finding themselves in the position where someone asks, “Which of our contracts has an auto renewal” or “Which of our vendor contracts has a built-in price increase in the future” or “How does a clause in this contract compare to what we typically agree to”?” Answering these questions usually requires someone going through the documents one by one. The task is extremely tedious, uninteresting and error prone. It’s one of the least satisfying things an employee can spend their time on.

Nomad Data’s Document chat can extract textual information across large document stores and convert the answers directly into a CSV file. This allows for simple extraction of information across these vendor agreements extremely quickly. They data can then be imported into another software platform or loaded right into Excel or an equivalent spreadsheet.

Regulatory Compliance

A large challenge facing many companies is complying with the ever-increasing volume of regulatory requirements. These requirements can vary by country, by state, city or vertical, and to make matters worse, the regulations are constantly evolving, meaning a practice you were undertaking a year ago may now violate one of them. Several clients came to us and wanted to use Document Search to check their own work product, their internal policies or even their pricing strategies against these large volumes of regulations. The engine has primarily been used in two ways for this use case:

1. Notify someone as to what regulations have changed and how
2. Returning specific issues with various documents with respect to individual regulations

Research Aggregation

Companies are often sitting on large knowledge repositories in the form of documents. They might be internal reports, notes, external research or publicly available documents such as company filings or announcements. Manually identifying trends or specific passages across these large volumes is extremely time consuming, and after a certain number of documents, no longer feasible. Below are several examples in this use case which help people manage this information much more effectively:

1. Public Company Filings & Earnings Calls – Users are asking about specific trends across dozens or hundreds of filings & earnings call transcripts
2. Research Library Searching – Users that are buying or producing large research libraries are rolling out Document Search to allow employees to ask questions against these libraries and letting the AI do the heavy lifting of finding the right materials to answer the inquiries. These libraries often include employee notes from meetings
3. External Research Libraries – Several Nomad clients produce and sell these research libraries to their clients. They are building chat solutions based on Nomad Data to allow their own clients to quickly find information in these libraries, significantly increasing the value of their research

Due Diligence

Banks, Law Firms, Investment Managers and Corporates often find themselves in the position of receiving a deal room worth of information on an M&A target. These deal rooms may include hundreds or thousands of employment agreements, sales contracts, NDAs, vendor agreements and other important documents. The level of diligence that was possible in the past was extremely limited given the tight timeline of many deals and the volume of documents. With Nomad Data, clients are able to quickly:

1. Output a CSV with every customer contract, product name, revenue amount / type, etc and add them up to confirm that they match what a company is claiming
2. Check thousands of employment agreements for missing confidentiality or IP protection details
3. Review all sales contracts to find ones with non-recurring or custom work

E-Discovery

Law firms typically receive huge numbers of documents and emails during the course of a litigation or investigation. Before modern AI most of the search capabilities around these challenges were solved with more rigid, keyword-based approaches. With Document Search, lawyers and investigators can provide significantly more nuance around the information they are seeking. The system can not only return the relevant passages but also provide color around them, saving teams a significant amount of time.

Insurance Claim Processing

Insurance companies still have many manual processes around claim processing. While much is automated, there are still a large number of claims that cannot be handled by legacy systems. Nomad Data also supports document workflows that can take in multiple documents, apply an AI action against them and then push them into the next stage of the workflow. This can significantly reduce the human intervention in these previously human-curated steps.

These are just a handful of the recent use cases we’ve been seeing. The buzz around AI is palpable, but the challenge for most is figuring where they can get the most impact. It’s critical for the first several use cases to show significant return so more budget is unlocked to continue adding AI more deeply into a company. We will continue to post about new use cases we see emerging.