August is a fast-growing legal AI company focused on building end-to-end workflows for mid-sized law firms. It serves customers on four continents, including Hicksons Hunt & Hunt and Mission Australia.

We sat down with two of the cofounders, Thomas Bueler-Faudree and Rutvik Rau, to talk about starting August, why mid-sized firms are the real AI winners, competing with Harvey and Legora and the evolution of the billable hour.

Thomas: We founded August during our third year at Columbia. We had met freshman year in a campus research lab, and we worked together on numerous research and computer science projects. We saw how the AI models kept getting better, and we started looking at professional services and how AI could automate parts of that work.

I’d worked at law firms in Alaska. Rutvik had worked at tech companies. Joe, our third co-founder, had worked in professional services at Blackstone.

Rutvik: We were in college when AI was still in its infancy. It gave us a lot of time to understand how these models work, how they interact with the day-to-day tasks people are doing. Back in 2023, everything was experimental. There was no clarity on the real use cases of AI. Just two years later, you see AI touching every part of the industry.

Thomas: We felt that the legal profession had lots of manual and tedious tasks that could be made more efficient with AI. We started doing some market research, and spoke to more than 100 attorneys to understand their workflows from first principles, and where AI could come in and assist. It helped that our dormitories at Columbia were right next to the law school, so we sat in on lectures, took classes, and spoke to professors. We put a minimum viable product together and started showing it around.

One of our first customers was a trade organisation in Washington DC. It was a seven-figure deal. We didn’t have a platform yet, so we had to build it almost overnight.

From there, we essentially built most of the platform out of our dorm rooms. I remember coding late into the night.

Rutvik: Back then, there was a novelty to the models. Most lawyers hadn’t heard of OpenAI or Claude. We leaned heavily into that. We said: we understand AI extremely well; we don’t understand law as well. So, we wanted to find a partner to co-build with. And it truly was a partnership as well. We figured out workflows, interfaces, and tools together. That allowed us to be very customer-focused from the beginning.

To be honest, those first customers were just betting on us. They were betting on three Columbia students who studied Computer Science out of their dorm rooms, saying, “Maybe these guys know a little bit about AI. We know the law. We’ll figure the rest out.”

Thomas: At the same time, protests were happening at Columbia. We were trying to sell to law firms downtown while protests were happening right outside our dorms.

I remember going to a meeting with an Am Law 50 firm during that period. We just escaped campus just before the police moved in.

It was surreal – going from a dorm room to a boardroom in the same day.

Rutvik: We’re a second mover. We’ve learned from their successes and mistakes.

We came in without assumptions about what the right product experience should look like. That let us reach feature parity quickly, and in some cases, exceed it.

Thomas: Customers who’ve piloted August alongside Harvey and Legora have told us August outperforms both. That’s coming from them, not us.

Rutvik: Another benefit is that all three founders are technical. We’re product and engineering-focused. That’s what allowed us to build in six to eight months what took others two to three years.

We’re not just competing with Harvey and Legora. We’re also competing with horizontal model providers like OpenAI and Anthropic, who are increasingly building vertical-specific tools.

Our constant question is simple: How is August always better than Harvey, Legora, as well as ChatGPT?

It’s still early in this industry.

If you look at social networks, MySpace wasn’t the eventual winner. Facebook saw what worked, what didn’t, started in a very specific niche, and scaled from there.

Our niche is mid-sized firms. We’re building deeply for that vertical, iterating faster, and then expanding outward.

Thomas: I’ve heard that Harvey is a small engineering team attached to a 500-person sales team.

We’re the opposite. We have a small engineering team attached to a small legal team attached to a very tiny sales team.

Most of our team are engineers and lawyers working directly on the product and with customers. Engineers and lawyers sit side by side.

Engineers learn how legal work actually happens. Lawyers learn how AI systems work.

That allows us to build the kinds of tools lawyers will actually use for generations to come.

Thomas: We work very closely with firms to identify their most valuable workflows. We deliver the platform with those workflows built in, and lawyers can edit them for different practice areas. The product is extremely customisable to how lawyers and their law firms operate.

Rutvik: This reflects a broader shift in AI. The first wave converted labour into software. Now agents can do end-to-end work.

We deliver workflows that are 90 to 95 per cent complete. Firms then come in and customise the last mile.

Lawyers don’t start from scratch – they’re given packages.

That’s where legal AI is heading.

Rutvik: Every lawyer knows the traditional billing model will change. Maybe not tomorrow, but over the next five to ten years.

Lawyers want to be valued for judgment, expertise and perspective. Not just their time.

August gives firms end-to-end workflows, which let lawyers sit down with clients and rethink pricing models and ways to package work.

The biggest reason firms hesitate to adopt fixed fees is uncertainty. They don’t know how long the work will take. AI removes a lot of that uncertainty because it does the grunt work upfront.

Firms like Hicksons in Australia are already leading here. They’re working with clients and other firms to deliver end-to-end work at a structured price.

Thomas: I don’t think the billable hour ends – but it changes.

We work primarily with mid-sized firms across the US, Australia, India and elsewhere.

When we ask whether August saves time, the answer is usually yes. Sometimes it’s no. But what it consistently allows firms to do is bill for work they could never bill for before.

There’s work that might have taken 100 hours in the past that no client would have paid for. Take a private equity roll-up of a small business. Traditionally, full due diligence didn’t make economic sense – the diligence could cost more than the deal itself.

With August, firms can now do that work and bill for it, whether under an hourly model or a fixed fee. That allows firms to bill for tasks that they could never have done in the past.

Rutvik: I actually think billable hours may increase in some areas.

The billable hour becomes a measure of judgment and expertise, not execution. And clients are happy to pay for that.

When a merger hinges on regulatory clearance or strategic advice, clients will pay high rates for that judgment. What they don’t want to pay for anymore is manual execution.

So you end up with bifurcated fees. Higher billable rates for strategic work, and fixed or flat fees for execution.

Thomas: Mid-sized firms have the biggest opportunity to grow in this AI environment.

Large, multinational firms often buy tools like Harvey or Legora, but they don’t always see real operational value. In many cases, AI becomes a marketing story.

They also might use AI to justify rate increases, but it doesn’t fundamentally change how they operate. It’s not a real business expansion.

For mid-sized firms, AI expands what they can do. It lets them go downstream and upstream – taking on work they couldn’t justify before.

Mid-sized firms are also more aligned internally. They have partners who genuinely want to change how work gets done, and they don’t have large innovation teams slowing things down with multi-year implementation plans.

Rutvik: AI is still very new, and whenever people learn something new, there’s a natural apprehension. The firms doing this well are leaning in with both feet. They’re willing to experiment and make mistakes.

AI implementation isn’t one-size-fits-all. Every firm has its own nuances, which is why we lean so heavily into customisation, and why August is so easy to customise. What we always tell firms is to just get started and actually use the tools.

The biggest bottleneck we see is firms thinking they need the perfect plan before they begin.

The only way to understand what works is to try things, see what doesn’t, and figure out how lawyers get comfortable with the technology.

Thomas: We also see very clear usage patterns.

Adoption peaks with early associates and senior partners.  Early associates are AI natives - they already use these tools at home and at work. Senior partners, on the other hand, know their clients incredibly well. That context makes it easy for them to leverage AI and build workflows that actually matter.

Where firms struggle is with mid-career lawyers – junior associates, senior associates, and early partners who don’t have the time to learn new tools. They don’t yet have the leverage senior partners have, and they didn’t grow up using AI in law school.

For that group, support really matters. Firms need to invest in training them. Adoption has to be top-down and bottom-up at the same time.

We also see that exploration drives adoption.

Our data shows that when a user engages with more than three parts of August – for example, the web platform, tabular review and the Word plugin – daily active usage jumps to over 70 per cent. That’s comparable to tools like email or Slack.

At that point, people stop seeing August as a tool and start seeing it as their workspace – where they get work done every day.

That’s why we built August Academy.

We built it with partner firms and tracked usage before and after training. Before the Academy, lawyers used August for around 5 to 10 hours a week. After, that jumps to over 30 hours a week.

Thomas: Last year was our first raise. We raised US$7 million from NEA, Pear and angels from OpenAI.

Before that, we were bootstrapped. We focused on talking to law firms, getting our first customers and building the product.

Over time, it became clear there was a lot of potential here. We were still in college, and I knew I didn’t want to take a job at a big tech company, a bank, or go to law school.

As the customer base grew, it became obvious that if we wanted to do this properly, we needed capital to hire the best engineers and develop the product.

At that stage, fundraising became about scale. This isn’t something you can do slowly if you want to build a category-defining company.

Rutvik: Expansion – both in terms of product and geography.

We’re expanding into South America, opening offices in Sydney and Mumbai, and aiming for 35 jurisdictions by year-end.

We take a partnership-first approach.

In Australia, Hicksons became an anchor partner and helped us understand what Australian law firms need.

As we’re expanding, we want to build out our anchor partners when entering new jurisdictions.

That approach lets us move faster in new markets.

* This interview was edited for brevity and clarity.  This article is part of our editorial content and is not a paid promotion.