AI in Legal Tech: A Conversation with NetDocuments' Dan Hauck

How is AI changing the way users interact with analytics platforms and what they expect from them?

People now expect AI to be in all the applications they're using. They know how to go to generic places like ChatGPT to use it in their personal life or for work. Where we're excited is how we can make it easy for them to find AI opportunities when they're doing legal work in their documents.

Being in that workflow and having it be a natural extension, oftentimes parallel to the way they're already working. As an example, we've released a new editing tool that lives inside Microsoft Word that allows you to give it instruction just like you would to an associate: "Make it more buyer friendly" or "Change these clauses to reflect this sentiment."

Just as you would get a markup copy from an associate, now you can get it back from your assistant inside Word. Having that natural feel makes it much more accessible for lawyers.

How do you balance AI's power with user control and human judgment?

Our goal at NetDocuments is to help lawyers do their best work. There are three things that are important to us from the beginning of the AI journey, designed to cater to the concerns of lawyers as they adopt AI.

First, picking a model and making sure your data is always secure. We were very early in thinking about what happens to our customers' data and how we ensure it is protected. We got exemptions from our model providers so that if something is flagged for content review, that doesn't happen for our customers. We do that in every jurisdiction. We think about where your data is processed and we process in that region. We have models in all our different locations.

Second, what does the AI have access to? NetDocuments as your DMS already understands who has access to what documents and what matters. As you engage AI, it's not going beyond what you should have access to or what your ethical obligations require.

Third, when you can tie the documents you're working on to your questions, prompts, and workflows, the outcomes are not just better from a low hallucination standpoint, they're more on point. They sound more in your voice and reflect the advice you've given in the past.

When you build a playbook automatically based upon dozens of contracts you've already done inside NetDocuments, then when you apply that playbook on the next one, you can have confidence it will reflect the work you've done in the past. Using a secure and properly governed model, ensuring it has access to properly governed content, and grounding it with your data and experience.

How do you measure effectiveness and determine meaningful metrics?

In some cases, it's straightforward. For due diligence reviews where you're extracting information from subcontracts, we encourage customers to compare previous work they've done with other systems or manual review to running it through NetDocuments. That provides a good benchmark to see if they would have gotten the same or better results with a new AI-based tool.

We co-build with lawyers, taking a process with tedious and evaluative steps, breaking it down and building a workflow around it that uses AI. Then we give it to the lawyers and ask if it's as good or better than what they've used in the past.

We've worked with firms on complex transactional agreements, documents that are hundreds of pages long with 15 to 20 pages of instructions on how to evaluate the document, make suggestions, and identify issues. On the litigation side, we've worked with firms who like to do deposition summarizations in a certain way with specific formatting.

When you have a platform that can do those things, it gives lawyers comfort that this implementation of AI understands the way they work. They don't have to change much to use it. Once they've looked at it enough times, they start to build trust in the solution. We all have that learning journey where we understand where it provides value and what workflows are most conducive to applying generative AI.

How is AI creating opportunities for growth and experimentation?

We're always thinking about how to improve the lawyer's experience and make it possible for them to do their best work. AI provides an opportunity to do that.

We've built an assistant that can readily access documents you're working on and answer off-the-cuff questions like, "Looking at these 10 contracts, which are governed by New York law?" That's a common use case built right into the platform.

For more complex workflows, we have a workflow tool and an entire studio of pre-built apps. We have almost 20 studio apps available for customers to extract content from documents, build playbooks for contracts, summarize documents, and build timelines.

One of the most exciting areas is helping lawyers understand what's inside their NetDocuments so they can find it more easily and use that knowledge for an agent-ready document management system.

We've been investing heavily in two areas. First is AI profiling. For years document management has had profile attributes like what kind of document this is and what office the lawyer is in. Very basic attributes. We found there's an opportunity with our dynamic metadata to share much more and use AI to fill out that profile card.

Now we might say, "OK, this is a lease agreement and here are the 15 deal points of that." What's the term of the lease? What's the monthly rental rate? What is the address of the property? All of those things can now be stored directly alongside the document as metadata and populated with AI. That happens automatically just by saving the document into NetDocuments. We're working with firms to be able to do that across millions of documents inside of their repository, which then allows them to answer questions like, "How many trademark applications have we filed in Canada in the last six months?" That would have been a very difficult question to answer before. Now it's possible because of this AI profiling.

The second piece of augmenting the platform has been introducing semantic search so that now it's not just using keywords to find documents. It's being able to understand what's in the documents, to understand the concept and search for those concepts and bring in a wider and more accurate set of content to help lawyers find what they're looking for.

That's something we've been working on for a couple of years now to build into the platform. What we're finding is it not only helps lawyers, but it creates that agent-ready platform so that as our assistant or a tool from a third-party application needs to find something inside NetDocuments, it can use all of these tools to land on the right set of documents and to inform an AI-powered workflow.

Do you have a customer advisory board or some mechanism for customers to not just provide feedback, but play around with beta versions of potentially new features?

We do. We have a robust customer engagement program. First and foremost, through our customer success team. We have several customer success managers who are engaged with our clients and who routinely help manage those conversations, as well as different groups like our user advisory group, our customer advisory group and different things like that. It's always been part of the NetDocuments DNA to get that customer feedback and put it into the platform.

Now with AI, people are just excited. We're seeing more lawyers wanting to participate in our customer product discovery sessions. It's the first time I've seen this much interest from the lawyers. We know this is an exciting area and many of them want to be on the front lines of it. We're providing ways for them to do that.

For example, last week I was on-site in Washington, D.C. with another NetDocuments C-level executive, three directors, and several team members for three intensive, six-hour customer sessions. We’re heading to New York next week for similar engagements. These aren’t just feedback meetings, they’re immersive, collaborative conversations where customers actively influence our roadmap and engage with early-stage concepts. It’s a deliberate extension of our engagement strategy, and a reflection of how deeply we’re committed to building alongside our users.

If you're a NetDocuments client or considering becoming one, we invite you to learn more about our AI-driven innovations in legal tech. Visit our website or reach out to your account rep to get involved.

About Dan Hauck

Dan Hauck is a former practicing lawyer and is passionate about delivering enabling technology to the legal industry. At NetDocuments, he is the Chief Product Officer and responsible for all products and the user experience across the platform. Dan also develops the strategic vision and roadmap for the company. Previously, Dan was the founder and CEO of the ILTA-award-winning company ThreadKM, a knowledge management and messaging platform for professionals. ThreadKM was acquired by NetDocuments in November 2017. Dan was a 2024 Fastcase 50 honoree and has spoken at numerous events about legal technology, including ILTACON, Lawtech, and LexThink.