Eigen launches expanded IDP platform with GenAI

last updated:

Eigen launches expanded IDP platform with GenAI

by:
last updated:

Eigen 6 release shows how GenAI and expertly trained machine learning models are better together than apart

It is always good to catchup with Lewis Liu, CEO and co-founder of Eigen Technologies, and Tom Cahn, Chief of Staff and principal bottlewasher. They hosted us at the London office in late November to unveil the new Eigen 6 platform, and we caught up earlier this week to discuss how the launch is going. Before we go all “techie” on you, I bet you didn’t know that Lewis nearly became a professional abstract artist, instead of the founder of a successful AI software startup. Tom has an impressive Liu painting hanging on the wall behind his desk.

Background

Founded as a 3rd Wave IDP company back in 2015, Eigen originally focused on the financial services market to help financial professionals analyze the complex documents common in their sector. For example, when LIBOR rates change, Eigen helps currency traders to quickly review their contracts, some of which can run into the hundreds of pages. Eigen also helps financial and insurance companies to analyze asset risks and loan agreements. The company is named after a math equation that has special significance to the founders. Eigenvalues are a special set of scalars associated with a linear system of equations. Perhaps the best-known business use of eigenvalues is within Google’s page rank algorithm. Eigen has raised nearly $80 million to date over three rounds. Goldman Sachs, an Eigen customer, became an investor.

Eigen 6 release overview

While the financial services market is still an important business for Eigen, the company has expanded into other document-intensive use cases across industries. With the Eigen 6 release, the team has solidified the company’s ability to solve many other document problems such as invoices, bills of lading, IDs, tax forms and much more. Eigen has also strengthened its pre-processing and post-processing tools and can now offer an industry-standard, end-to-end IDP solution.

On the GenAI front, Eigen 6 uses LLMs where it makes the most sense. As we have reported time and again, GenAI alone is not an IDP solution. The following slide from the team shows where GenAI adds the most value, and where Eigen’s own AI models are the best choice. Note also how both work together at several points.

Eigen also introduced Generative Insights. The LLM adds the capability to extract inferred answers from a document, something that Eigen’s ML models alone don’t do. The following screen shot shows how the user can interrogate a quarterly report and ask a complex reasoning question like, what is the Debt to Asset Ratio? Working on the data carefully extracted by the Eigen ML model, the LLM is very good at answering this type of question.

The team also showed us how Eigen 6 can tweak the confidence scoring knobs across data fields, and then intelligently assign the most cost-effective AI model to each task. We think this could be a significant competitive advantage for Eigen. LLM tokens are expensive compared to discriminative AI models. The team gave an example of the large cost difference: for one loan document, it could cost $2.17 to process one question in OpenAI’s GPT 3.5, compared to just a fraction of a penny using an Eigen model. So, the ability to choose the best option has immediate business value. We suggested that Eigen should build a simple cost analysis calculator on top of this; CFOs will love it.

February update

At the February briefing, we asked Eigen, what’s been the customer feedback so far on using the GenAI features in Eigen 6? It’s still too early to tell. The answer was not unexpected as we also hear this from other IDP vendors. What the Eigen team are seeing at the early stage is that GenAI turned out to be both less accurate for data extraction and far more expensive to use, when compared against Eigen ML models trained for specific document types. However, users are finding GenAI adds value at several stages: model training, document classification, multimodal extraction, summarization, and post-processing steps such as data transformation and validation.

Eigen also shared what we thought was a very interesting anecdote about the impact of ChatGPT on market awareness of AI. The sales and marketing team talks to hundreds of potential customers (e.g., prospects or leads) each month. Before the ChatGPT earthquake, the team had to spend most of the conversation explaining how AI works and why a business should invest in it. Most businesses were slow to decide and implement. Now, prospects come to these calls already having a basic knowledge of generative AI and are more likely to move faster to implementation. Many of these prospects have been tasked by their company with finding GenAI solutions.

Summary

With Eigen 6, the company skillfully added GenAI alongside its own AI models and is poised to compete on level ground with other end-to-end IDP platform providers. Eigen’s proven expertise at managing complex unstructured documents will be a distinctive competitive advantage in some use cases, as most of their competitors started with structured and semi-structured documents and are just now moving into unstructured.

Leave a Comment

We use cookies to ensure that we give you the best experience on our website. If you continue to use this site we will assume that you are happy with it.

Work Intelligence Market Analysis 2024-2029