AI in Law: Six Ways Axiom is Riding the AI/LLM Wave (and how you can, too)
By Chris Frickland
All we see in the news lately is AI this, AI that. The legal landscape is shifting more than ever before as companies across the globe grapple with novel implications and unprecedented complications. At Axiom, we’re no different. But we’re embracing this wave. Here’s how we’re adapting and learning to use AI in law, with some tips for you, too.
As head of Technical Product Management, I help build the technology Axiom needs to enable us to find the right lawyers for our clients. Over the past two years, I’ve expanded my areas of focus to include artificial intelligence (AI). I head up our AI research to understand how we can use the science and AI models to make the job easier for our internal teams, but also for our clients and the legal talent with whom we work.
A typical day for me involves talking to our Science Team to evaluate the results of models that we've built and discuss some of the explorations that we've done. I also work closely with our Talent Success Team - the team that's sourcing the amazing talent for our clients - about how they're using these models, how they're interacting with them, what's working, what's not, how we can make them better, and what we need to improve upon to make their jobs easier.
At Axiom, we use AI and machine learning (ML) together as an all-encompassing toolset: something that can help us find the right talent, at the right time, and get our clients the help they need in the most efficient way. We believe that ML helps augment a lot of what we're doing, and AI, specifically generative AI, can have really useful applications when done thoughtfully. A lot of similar players in our field have come out with tools too quickly, prioritizing speed over customer experience – meaning the precision of a lot of these tools is subpar. Axiom is taking a more thoughtful approach and being selective about when and where we use this science.
To understand how we put our AI strategies to use in practice, we use both traditional machine learning and generative artificial intelligence combined strategically that works well for us in a way that’s practical, pragmatic, and cost-effective.
Six considerations in adopting generative AI:
1. Use more for the business, not the practice.
In general, we try to be pragmatic about how we're implementing the generative AI side and use it only when it makes sense both from a risk and benefit perspective (remember, we are in the legal industry, so this is always top of mind). In practice, we use it for things like summarizing phone calls, aiding sales training, and generating job descriptions. We don’t use it to read or write contracts, as that requires a high degree of precision, something that is not feasible with large language models (LLMs) currently. For applications within our legal teams, we are looking at copilot tools and assessing them as they come out. Some of these may help teams with legal research and document review, but we will maintain a human in the loop – we aren’t looking for tools that would replace a lawyer’s legal judgement or due diligence.
2. Forecast your supply and demand.
Our first step is ML running a program that we use to forecast our supply and demand. Essentially, this means we're running models to make sure we're finding the right person for our client at the right time, ensuring that we have our bench built up, predicting the demand our clients are coming in with, and that we hire those people into our bench. So, when a client comes to us and says, “I'm looking for a lawyer in real estate specific to Atlanta,” we've already forecasted that and checked our bench to ensure we have great talent already ready to go. This saves on time spent matching and improves our relationships with clients.
On the flip side, we also use this to avoid over-hiring in certain areas, making sure that if we believe demand is going to peak and start to go down, we don't continue to hire in that area and build up talent that is waiting for work. This way our lawyers won’t be waiting for an engagement that may not appear for some time. It’s a balancing act to keep our work flowing steadily without clients or talent kept waiting for engagements to begin, and the help of ML makes this balance successful.
3. Transcribe exploratory calls to create job descriptions.
At Axiom, we have a massive bench of legal talent, consisting of more than 13,000 lawyers and legal professionals all around the globe. When our sales team works with a client on finding out what lawyer or project support is needed, it used to be overwhelming to search for the perfect fit in an efficient way. Now, we combine call transcriptions with job description creation using generative AI technology.
When a client has a call with a sales team member, we take the call transcript from our call recording service and use it to generate a job description our talent sourcing team can immediately take to our bench of lawyers to find the right person for the job. We use AI here to take the summarization of that phone call to pull out those exact keywords from the conversation to see what the job description should look like. Once we have a job description built using these keywords, we can start matching it to talent resumes that also include the same keywords, and almost with the snap of our fingers, we can start to highlight who would be a good fit for that particular job.
We also take the generated description and allow our bench of talent to seek out what’s the best fit for them. We share that job to our bench using our internal talent platform, providing our talent the opportunity to self-identify when they are the perfect fit.
Similar to how resumes only offer a snapshot of lawyers’ work experience and may not be totally comprehensive of their abilities, we found that showing job descriptions to our talent and letting them actually raise a hand and say, “I’d be an amazing fit at this job, I just don’t have that specific thing you are looking for listed on my resume,” helps us expand what we know about our talent and is something that AI can’t help with. Ultimately, it's about getting the right legal talent to clients faster.
4. Use AI to match the right job to the person.
We use a more traditional machine learning model to match that job description with talent by searching the tens of thousands of resumes and saying, “These five people would be a really good fit; you should take a look at them.”
We are using our deep understanding of talent and what our clients look for to help kickstart our internal searches, offering our internal sourcing teams a one-click shortlist for their opportunities and helping find the proverbial needle in the haystack. This shortens the time between when a client needs help and when we can show them the available lawyers matching their exact needs.
In addition, this helps us find talent that might not be top of mind - either new talent or talent that is just becoming available - as the ML can search our vast database of people and find the best fit.
One of the really interesting benefits of AI is that it can provide an instant look across our entire database of talent. It doesn't matter who you talked to recently and it doesn't matter who we just hired to our waitlist; this model cares more about the best fit based on what the client described the work as being. This eliminates the potential bias of leaning toward a lawyer with whom you just got off the phone. It instead looks at what the client is saying they’re looking for and the technology then compares that with the thousands of available talent on our bench to make a recommendation.
For example, a lawyer might have a four- or five-page resume with a key experience a human may miss when reviewing the thousands of potential options. Now, picture a client needing counsel support for a specific AI-related project and a lawyer with the exact experience needed extracted from the long resume pops right up. Who knows? You might have been missing your perfect fit all this time, simply because there were too many resumes to read.
5. Don’t forget the human touch.
By combining this tri-sourcing technique, consisting of the opportunity feed, artificial intelligence, and our (human) white glove service we’re known for, our clients get the best results. Each of these avenues has a different speed and a different reason.
Our two main goals in incorporating AI are speed and quality, but we know that the latter is more difficult. Finding the perfect lawyer will take some time. So, we are focusing on speed by using the initial responses to aid our team in using that signal to better enhance their human-powered research for quality talent. As we continue getting more feedback through this process, over time our AI will learn the key details for which we’d like it to search, which will help us get amazing quality with already impressive speed.
6. Continue learning, improving, and looking for the next steps in your legal tech journey.
We hope to move more of our internal AI and ML strategies into external cases like this in the future. As we begin to weave our machine learning into our searching and finding legal talent, we want to begin using more GenAI alongside it to help explain why we are finding these people.
The next evolution of our AI journey is finding out how to help inform our clients and internal teams about why this or that lawyer was a good fit, or what was unique to that person that stood out for this job. We’re taking the job description and the lawyer’s bio and trying to find those missing layers so everyone can quickly understand why they are seeing these recommendations.
Currently, our clients can access the first glimpse of our modeling through our free digital platform, Access Legal Talent, where you can peruse the deep bench of legal talent that Axiom has and browse similar talent, all powered by ML. I encourage anyone to visit this platform and find the perfect lawyer you are looking for today. These tools have the potential to transform the way legal services and legal work are delivered, and Axiom is ready to embrace the future.
If you’d like to hear more about how Axiom specifically employs the use of AI and LLMs, watch the latest installment of the Higher Bar, where I join Zach Abramowitz, CEO of Killer Whale Strategies and LegalTech expert, and we discuss how these powerful AI tools are changing the way we work. This discussion will equip you with a better understanding of how LLMs are likely to impact not only the way you practice, but also how you can actually leverage them effectively in your practice, law firm, or legal profession.
💡 Learn how to better manage the unique challenges and complexities presented when onboarding, utilizing, or developing artificial intelligence.
Chris heads the Technical Program Management and Data Science functions at Axiom. As an inaugural member of the Axiom Research and Development Team in 2018, Chris has delivered all Axiom’s platform initiatives, spanning internal tools for matching Axiom’s global legal talent to its prestigious clients, a digital experience for Axiom legal talent to manage their end-to-end experience with Axiom, and Axiom’s first push to present its black book of legal talent on the web to current and prospective clients. In addition, Chris spearheaded Axiom’s introduction of machine learning technology, pioneering the use of Axiom’s vast data sets built from years of manual talent-to-client matching to recommend new and unknown talent to Axiom’s clients. The resulting “Magic Lawyer Finder" technology aided Axiom’s record 2021 growth in an extraordinarily talent supply-constrained environment. Prior to joining Axiom, Chris held technical leadership positions at Tableau, Starbucks, and BBI Engineering, enabling him to dive deep into technical challenges in museum-quality technology installations, city-scale infrastructure, and enterprise-scale data analysis.
AI in law is changing the legal landscape as companies & law firms grapple with using AI. Here's three questions to understand AI best practices.
Axiom Lawyer Nancy Siegel brings over 20 years of combined experience in the life sciences and insurance regulatory industries to her work with global pharmaceutical companies.
Law firms are increasingly collaborating with alternative legal service providers to win and retain clients.