Show Notes 190: The Future of Legal Tech: Why AI is Democratising Contracts, and What It Means for Lawyers

In the latest episode of the Cambridge Tech Podcast, hosts James Parton and Faye Holland chat with Rafi Farouk, co-founder of Genie AI, about how artificial intelligence is fundamentally transforming the legal industry. Spoiler alert: it's not about replacing lawyers, it's about reimagining their role entirely.

The Weekly Tech Round-Up

Before diving into the main interview, the hosts covered some exciting Cambridge tech developments:

• Overt STEM Summit launched to accelerate inclusive innovation across the Oxford-Cambridge growth corridor, with 22 award categories celebrating excellence in STEM

• Cambridge Science and Technology Torchbearer Award introduced by Business Weekly, honouring companies where commercial success and societal value are inseparable

• CUSP AI partnered with Nvidia to revolutionise molecular simulation for drug discovery and material science

• Kigan raised £10 million to strengthen cybersecurity resilience in the EU and US markets

• Start In Cambridge unveiled a one-stop shop helping founders navigate UK company formation, taxation, legals, banking, and community connections

Meet Rafi Farouk: From Trading Floors to Legal AI

Rafi's journey is the stuff startup folklore is made of. After studying philosophy and economics at university, he became a trader managing over £1 billion at just 22 years old (yes, including Vatican investments). But craving something with genuine social impact, he pivoted to machine learning and founded Genie AI eight years ago, before GPT-2 even existed.

Key insights from Rafi:

"We're democratizing expertise and what we're really selling is trust."

"The bigger market is all the end businesses, and we feel we're making a bigger difference in the world by serving that mid-market business rather than the big players."

The Technology: Eidetic Intelligence

Genie's secret sauce is a patent-pending architecture called Eidetic Intelligence, which dramatically outperforms ChatGPT and Claude for legal tasks:

• 86% accuracy on benchmarks vs. GPT's 48% a significant leap in legal quality

• Handles unlimited context length across 20-30+ documents simultaneously

• Creates a knowledge graph of company policies, templates, and negotiation behaviour

• Maintains legal quality through intelligent compression and gating mechanisms

This matters because most law firms operate on a template and tweak model, charging tens of thousands for slightly modified contracts. Genie flips the script entirely.

The Jobs Question (And Why It Matters)

Let's address the elephant in the room: yes, jobs will be lost in law firms. But Rafi isn't dodging the question.

"I do think jobs will be lost in law firms and we shouldn't beat around the bush with that."

However, Genie is partnering with Lord Wahey and Maker Life to create "Trust Engineers"—lawyers reskilled to set AI rules, standards, and guardrails. The company is publishing this training program publicly later this year.

The Bigger Picture: The Agent Economy

Perhaps most fascinating is Rafi's vision for the future of work itself. As AI agents handle more tasks autonomously, company structures are already shifting:

"Everyone is being an autonomous end-to-end team. Everyone's doing the work end to end via agents."

Rather than cross-functional teams slowing things down with collaboration, we're moving towards generalists managing multiple AI agents—essentially, one-person companies at scale.

Why You Should Listen

Whether you're a founder thinking about your next tech stack, a VC evaluating legal tech investments, or simply curious about how AI is reshaping professional services, this episode delivers genuine insight without the hype.

Genie AI is already backed by Google Ventures and Coaster Ventures, operates across 35+ countries, and serves everyone from SMEs to enterprises. But the real story? It's about reimagining an entire industry.

Tune in to hear the full conversation on the Cambridge Tech Podcast—available on all major platforms.

Now you can get directly involved with the show. Leave us up to 60 seconds of audio by visiting CambridgeTechPodcast.com on your computer or phone and click Call the Show from the menu. Send us your news items, events to promote, Leave us feedback or suggest a guest.

AI generated transcript

James

Welcome to the Cambridge tech podcast, Talking all things technology from the heart of the UK's tech capital. Here are your hosts, Faye Holland and James Parton. Hi, James and Fay. This is Andrew Hatcher with a quick update on a new event at the Bradfield in Cambridge on May 13th. We're launching the very first AI for Business event and it is designed specifically for SMEs who want to actually use AI, not just hear about how clever it all is. So what makes this difference is that we are focusing on what works in the real world, using other SMEs to show you how to. We'll be sharing a live case study from a marketing agency that has implemented AI, including what they tried, what didn't work and what actually delivered results. And then we'll go a step further.

James

We'll introduce you to at least three practical demos of AI tools that you can start using this week across areas like sales and tenders and lead conversion. So you won't just leave inspired, you'll leave with things you can actually implement immediately. So if you are curious about how AI can genuinely help your business, sign up on Eventbrite. It would be great to see you there. Thanks very much.

Faye

I'm Faye.

James

And I'm James. Faye, what have you been up to?

Faye

It's a happy new month, by the way, so no pinches and crunches because it is the fifth, but let's do it virtually because I'm nice like that last week. Good week, really busy. It was again, one of those weeks I did this, the networking session at Impulse that I think listeners will remember. Every year I rock up at Impulse and teach the next generation of entrepreneurs what to do with regards to networking. And I just love that program. I think it's pretty much the only entrepreneurial program that I'm still supporting, but it's just got such a lovely vibe. So I did that and then I went straight to the Business Weekly Killer 50 and we talked about that on last week's news. Martin Frost, one of the founders of CMR Surgical, was the keynote speaker. He gave a really authentic presentation.

Faye

Nothing was held back from that conversation. He talked very much about how they'd moved fast and they broke things to get them where they needed to be. He gave huge credit to the importance of hiring the best marketer, which was just music to my ears. And it was really good to see Patrick Pordage, to get the recognition in front of everyone that was in that room. So that was good. But the bit that made everyone laugh was this might be showing My age certainly is. He talked about the Morecambe and Wise sketch and aligned it to running a startup. So this is the one where they're sitting at the piano and one of them goes, okay, you know, that's just. It's just not right. And he said, no, I'm playing all the right notes, but not necessarily in the right order.

Faye

And everyone laughed because that is what the world of startups like. I know you're going to come on and talk about that in a moment. Anyway, so quite a nice little segue into your days. And then, Yeah, I spent three days in Manchester, my longest stint. I tell you what, I walk loads in Manchester.

James

Do you? It's a big city.

Faye

I know, but literally everyone walks everywhere, so I'm really getting my steps in. I think there's all of these other benefits to working on such a cool project. Anyway, so I had my first speaking gig, so that was interesting. I was on the stage talking about the partnership, so that was good. Up at Sister, which is a little bit like the Bradfield Centre, it's. I feel, my second, the spiritual home. So it was really great to be there and just meeting loads of people. And I'll put the link to the report that was released, it was interesting. It was all about regional engines for national growth. So I'll put it up on the link as well and then everyone can take a look at it. And then, yeah, just loads and loads of work. We did our last two partnership roundtables we had.

Faye

The two last sectors couldn't have been further apart. They were defence and creative industries. But there are actually opportunities that are surfacing that can bring both of those together. So it's really interesting to see how creative we can be with the partnership as well. And then it was lovely to have a nice long weekend. That was me. So what about you?

James

Two main things, I think, from last week. One is I went to the CW International Conference at Higgston hall, which was really nice to catch up with lots of friendly faces. Jen Richardson, Raoul from Anthro Tech, who you will no doubt remember from our interview with them nearly a year ago. Can you believe it's been that long? So they're. They've got some really impressive contracts coming into place, so we should bring them back on, actually, for an update, I think in terms of the conference content, I really enjoyed the keynote from Peter Hague, who's the Deputy CTO of the national center for Cybersecurity. So he was talking about how AI enabled threats are changing the world.

James

The way that Telcos are going to need to manage their networks, especially after Anthropic's Claude Mythos became the first AI model to successfully complete all the 32 steps necessary for a successful capture the flag penetration test. Yeah, that's scary, but I think awareness and planning is the key thing there. Right. Also really interesting point that because AI is so much better and faster at finding and fixing vulnerabilities, that has a knock on effect to the way businesses can need to improve their patch management as well without sacrificing stability in their systems. So lots for the telco crowd to think about.

Faye

And James, just before you go on, that whole sense conversation kind of links to what I was saying as well. And I think there's almost like you said, it's a little bit scary and let's use technology and try and make it in the most positive way. But I think there's almost like a, it's got an identity crisis at the moment because people don't want to talk about it, but they know it needs to happen. And I think actually we should probably do an episode on what it actually means and to do what we need to do for security, but also make sure we're doing the right things. So maybe that's something we can dig a little bit deeper on.

James

Yeah, and it was theme of the conference as well that where the military world and the civilian world used to be divided, there's now a kind of a collaborative attitude where the Ministry of Defense is looking for civilian collaborations, for example, in sharing of network infrastructure and access. Whereas previously the military communication networks were completely locked down, now they're looking at how they might use things like 5G, 6G, low orbit satellite networks and all these kinds of things just to enhance the propositions. So yeah, there's definitely some ground to cover there. My, my secret project is now not secret anymore. So we launched Launching Cambridge at the conference. So what is it? It's really designed to be a one stop shop to help founders set up in Cambridge.

James

And it's really took inspiration the conversations I had with founders over the years on their experiences of setting up in Cambridge. In parallel to the pressure of growing their startup, they have the additional pressure of thinking about UK company structure and formation, taxation, accountancy, legals, banking, fundraising, workspace, social and business community connections, finding somewhere to live, somewhere to send their kids to school. It's obviously really overwhelming. So this proposition is really just designed to try to help and address that. And I'm really pleased of the group of local experts that we've brought together. They really are just like the stream of the crop in Cambridge. So it's us at the Bradfield Centre, it's Mills and Reeve, it's Cambridge and Peterborough combined Authority, it's pem, Barclays, Eagle Labs moved to Cambridge and Cambridge Wireless are the initial launch collaborators.

James

So I'm hoping that's going to really make life just a lot easier for startup founders looking at Cambridge as an opportunity. So if you'd like to learn more or are interested in joining the group as a service provider, just ping me on LinkedIn or visit bradfieldcentre.com launchcambridge no, that's brilliant.

Faye

Just that sheer list of things about all the pressures of starting a company that you listed, that in itself is the reason to have new services and new options for people to help them land a little bit easily. So brilliant. And yeah, let's keep a watching eye on that and see what we can do. Maybe we should do one of those as an episode too.

James

Yeah, for sure. Okay. I think that's everything from us, isn't it, in terms of updates. So should we move on to the news? First story this week is Overt stem, which is dedicated to accelerating inclusive innovation across the Oxford to Cambridge growth corridor, has been launched. The Overt STEM Summit is where STEM community leaders come together to collaborate. It's conceived as a multi zone festival of ideas and brings together an innovation showcase, a pitch arena, academia led collaboration spaces and high impact learning exchange. All designed to remove barriers, accelerate partnerships and turn conversation into action. In addition, the Overt STEM Awards celebrate the people, the teams, the organizations that are driving STEM forward across the growth corridor. With 22 categories spanning education, innovation, inclusion and leadership, the awards spotlight excellence and inspire the next generation of STEM leaders.

James

Overt STEM's founder Mia Chandler says Overt STEM exists because the talent is everywhere, but opportunity isn't. We're building visible pathways and tangible opportunities so people across the growth corridor can, regardless of their background, can thrive and lead. In stem. This isn't just an event, it's a movement. So that sounds a very important thing to publicise.

Faye

It does. It's a lot of movement for starting here in Cambridge, aren't they? Next of all on awards, there is a new Cambridge Science and Technology Torchbearer Award that has been introduced to the annual Business Weekly award. And this award will honour a company whose commercial success is inseparable from measurable societal value. Delivered by Innovate Cambridge in collaboration with Cambridge Innovation Capital, the award aims to encourage a next generation, a next wave of startups and scaling companies who embed societal value into their business model from the outset. So Catherine Chapman, who is Innovate Cambridge executive director, said, we're proud to introduce this new award as part of the Business Weekly Awards, recognizing a new generation of technology companies redefining what successful innovation looks like.

Faye

Cambridge has long led the world in scientific and technological breakthroughs that have changed how we think and how we live. This award highlights those companies showing that innovation can and should strengthen the communities in which it operates. And you can apply for that award by finding out more information on the Innovate Cambridge website.

James

Great. Our next story is CUSP AI? That interview with CUSP is coming up very soon, so stay tuned for that. They've notched another major coup through an alliance with global giant Nvidia to address molecular simulation. Now, apparently molecular simulators are the backbone of drug discovery and material science, but they're notoriously complex. The pharmaceutical industry, for example, spends 15% of its revenue on R and D and more than 20% of the total R and D spend across all other sectors combined. And molecular dynamics is one of the heaviest computational lines in that budget. Ian Buck, vice president of hyperscale and HPC Computing at Nvidia, said, we've been working closely with the CUSP AI team to help push the boundaries of GPU accelerated molecular simulation. So congratulations to Cusp and we can't wait to bring you that interview with Debbie.

Faye

And our last piece of news is from Cambridge based tech company Kigan, who has raised 10 million pounds to accelerate the next phase of growth in the EU and US as enterprises strengthen cybersecurity resilience for the era of new regulation and investment in AI. Vincent Costange, CEO of Kigan, added, we're excited to join the high Caliber portfolio of companies backed by Celica. This investment allows us to accelerate our next phase of growth with confidence. Our customers are looking for practical ways to strengthen cyber resilience, simplify regulatory readiness and manage secure connectivity over the long term. This funding helps us move faster to support them with trusted future ready solutions.

James

That's the end of this week's news, so let's now move today's episode.

Faye

So today we're joined by Rafi Farouk, co founder of Genie AI, which is a company using AI to transform how legal contracts are created and negotiated. So Rafi, is great to have you on Cambridge Tech podcast with us.

Speaker 3

Great to be here. Thank you Faye.

James

So Rafi, we always like to start by getting to know our guests a little bit better. Could you Walk us through your background, your early kind of path, how you found your way into tech and entrepreneurship.

Speaker 3

Sure. So I've been a lifelong programmer the age of 13 or 14, creating websites in HTML and PHP. But at university I wanted to learn more about the world. So I read philosophy in economics to get a broad brushstroke on the world, then did what all my peers were doing and became a trader. Enjoyed the dark side portfolio of over a billion pounds at the age of 22, which is quite a lot of responsibility quite early. Traded with the Vatican, believe it or not. But then I wanted to get back to doing something that A had a tech focus and B, had a positive social impact. And so that's why, after running a further business, I then went back to a master's in machine learning. Lucky enough to be taught by google DeepMind at the time.

Speaker 3

Wrote my master's thesis in generative AI 9 years ago and then started GENIE off the back of that master thesis. And yeah, genie's mission is to empower everyone to agree with confidence when it comes to legal contracts.

James

So fail, jump in and talk more specifically about the product. That aha moment. Did you have an experience with the law that kind of triggered the or spotted the need or the gap in the market? Did you have the technology idea and you were looking for a problem to solve? Which way round did that come?

Speaker 3

Yeah, were the wrong way around. So we had the tech. There's a video of me eight years ago speaking about how generative AI can generate text and contracts, legal contracts, end to end. And that was before GPT2, I believe, maybe even before GPT1, before the current language model explosion. I believe were the first generative AI company in legal, but it was a tech insight and then we obviously had to speak to hundreds of lawyers and the industry to really validate the problem. But what that inspired me is the fact that law firms tend to sit on all these past contracts, all the past knowledge, and for every new deal they just. Not just, but they make a few tweaks to these template contracts and they're selling it for tens of thousands of pounds.

Speaker 3

And we just thought, this is ridiculous, we need to create some kind of legal brain that can read way more knowledge than the law firm and produce the result way cheaper and way faster.

Faye

So the idea then is to reduce the reliance on lawyers by using the technology.

Speaker 3

Pretty much it's democratizing expertise and what we're really selling is trust. Now anyone can go on GPT and ChatGPT and ask for a Legal contract. But we recently released some benchmarks showing that GENIE outperforms GPT and Claude on public benchmark that we've released by quite a lot. On one benchmark, GPT got around 48% and Genie got around 86%. So that's the kind of difference in legal quality you're getting.

Faye

And just delve a little bit deeper then into the effect what you're doing is going to have on the legal industry.

Speaker 3

Yeah, I think the big question is always jobs and I do think jobs are going to be lost in law firms and we shouldn't beat around the bush with that. I was lucky enough to be on Sky News two years ago talking about this, and the presenter asked me three times in a row, what's going to happen to jobs? What's going to happen to jobs? And back then I was a bit more. A bit more vague, but now I'm perhaps more blunt and I do think jobs will be lost. That said, we're also partnering with Lord Way at the House of Lords and a company called Maker Life to create a training program for lawyers to upskill and retrain into AI lawyers, which we're calling trust engineers, trust architects, people who set the rules and standards for the AI to operate automatically.

Speaker 3

So there is still a role for lawyers. It just will change. And that's what we're trying to enable.

Faye

And I think that's the important bit, because go back a few years, everyone was talking about, oh, we won't need lawyers anymore, but we are still using lawyers to the same extent. So it is that transition, but all technology is there to redefine jobs. I think it's really interesting that you're already looking at tackling that as well.

Speaker 3

I moved into legal AI to try and do something good and positive for society compared to perhaps trading. And I do worry sometimes whether the AI boom is net good. And so this is an initiative to offset maybe some of the job losses. And we're publishing this training program on the GENIE website, hopefully later this year, which will be publicly available. And it's not something we need to do, but it's something that we want to do for the benefit of society.

James

Can I just ask a supplementary question to that before we move on to maybe delving a little bit deeper into the tech? It's fascinating to hear you describe that, because I'm guessing in the early days of the company you had a decision to make, because it sounds like you could have built a tool for legal professionals that just does a better job of unlocking all of that insight and they could have used that in house to improve productivity. But you chose a path to go and compete directly with the legal profession. So talk us through that kind of choice and why you took the latter route.

Speaker 3

Yeah, it's a really interesting question. In the early days we did work with the big law firms, mostly to gain knowledge, understand the market, but also due to pressure from investors. But again, I left big banking not to go into big law, but to try and do something that changes the world. So we quickly pivoted to serving the end customer and it is harder. Many of our competitors have gotten quite big. You may have heard of Harvey AI Ligura by serving the big law firms, but in the end, the bigger markets is actually all the end businesses and we feel we're making a bigger difference in the world by serving that mid market business rather than the big players.

James

Okay, that's admirable. So you're obviously working at this intersection of law and AI and we've had a lot of AI startups on the podcast, as you'd expect. So let's get into the weeds a little bit. How exactly are you using AI to really drive the proposition forward?

Speaker 3

There's actually so many ways. It's a hard one to boil down, but if I try to categorize the key areas, the first thing I'll say is we just released a patent pending architecture called Eidetic Intelligence. And it's this type of intelligence that's outperforming CLAUDE and GPT significantly for legal tasks. We can essentially handle big multi document legal transactions. Imagine if you're building, if you're a construction company building a commercial development, you need to have 20 to 30 subcontractors. There's probably a main contractor, there's lots of different parties involved. Now if you put 30 documents into Claude or GPT and your policies, your templates and all your company history, a hundred other documents, chlord GPT will fail. It can't handle that context length.

Speaker 3

What we've done is we built this architecture called Eidetic Intelligence which memorizes all your past knowledge and information and in doing so it personalizes the outputs. But it can also operate over unlimited context length. So it does that by compressing the information into a knowledge graph and having a legal quality gating so that every decision is high quality despite the compression into a knowledge graph. It's probably a lot of words, but in short, GENIE can operate over unlimited context length, unlimited documents and project size whilst maintaining a really high level of legal quality.

James

How have you trained the model to perform like that?

Speaker 3

Yeah. So that's pretty much the creation of this knowledge graph plus legal quality gating. And it's something we've just submitted a patent for as well. But we essentially anonymize your company knowledge documents, policies, how you tend to create legal contracts in your business, your typical risk thresholds, your negotiation behavior when talking to clients. And then we distill all of that into a knowledge graph that can be used to train the models at test time. And yeah, this creates a highly personalized output which in turn enables quality.

James

Technology moves fast. And now so can you. Discover polestar, the all electric performance brand vehicles redefining how we drive precision engineering, minimalist design and software that evolves with you. Experience it at Holden Group in Norwich and Bury St Edmunds or let us bring the test drive to your door. Proud sponsors of the Cambridge tech podcast Polestar Electric performance redefined. Discoverldongroup.co.uk Polestar. You describe their complex structures and ingesting history. Does that imply that you're targeting more of a corporate enterprise customer or are you agnostic to the kind of people that could you use this in buying a house or doing a more kind of an individual kind of legal requirement like a will or something like that?

Speaker 3

Yes, you can use it for individual requirements and we do get many people doing that. We have quite a generous free tier again to allow individuals to use the platform as a benefit to society. That said, it's not our target market. Our main target is enterprises, mid market businesses and typically those companies that are more project based. If you're creating a construction project, an energy power plant, if you're consultancy, then you'll have more contracts and more of a variety of contracts to do.

James

Okay, and I guess my final question before I hand back to Faye and you must get this a million times, like how do you build the trust in relying on the AI to perform these tasks that you typically be paying a lawyer to do?

Speaker 3

Yeah, I'm glad you asked. And you're absolutely right. This is the biggest question this and the question of does is the market especially lawyers, afraid of what you're doing? Two biggest questions we often get. But on the idea of trust, it's pretty much releasing public benchmarks on our quality and we've just released this. Anyone can test our system against the benchmarks, anyone can test their system against our benchmarks and you can compare against GPT and Claude. And like I say, we're performing sort of 40 to 50% better than GPT and Claude. We believe this is the best in the market for legal quality on complex legal tasks.

Faye

Thanks. That's really interesting. Just going through a little bit about the sector and what you're doing. So what I would like to do now, Rafi, is talk a little bit about the company journey. So what's happened since you founded, how you found team development, those types of things?

Speaker 3

Yeah, so we're about 35 people now. That said, and I guess this is all relevant to the world now, but the unit of work has been decoupled from the amount of people. So actually smaller companies are in many ways performing a lot better than big companies with the use of agents and so on. But yeah, it's been eight years of running the company. We've between series A and series B. Very lucky to have investment from Google Ventures, Coaster Ventures, code of one of the first investors in OpenAI. So yeah, it's been a journey and yeah, lots of ups and downs.

Faye

So your comment there was interesting on employees and the model actually changing. Do you see a future where you're going to have that split between agents, humans? How do you know? Do you have any kind of forecast on how you would handle that?

Speaker 3

Yeah, we're already seeing it right now. Our company structure, our job roles have completely changed. And when people's job roles change, that is the thing that is most emotional and most challenging typically. So it's something that we've really carefully monitored and communicated about. Essentially what's happening is that before you need cross functional teams to deliver a service or a product. So you might have an engineer, a product manager, a designer, a marketer, but now you have one person that does everything. Because where businesses slow down is where you need to collaborate. Which sounds sad because I'm advocating for less collaboration, but that's the reality of where teams slow down in terms of organizational efficiency, where you need discussion, interfacing. Yeah, things slow down the fastest teams are teams of one. And what's happening now is everyone is being an autonomous end to end team.

Speaker 3

Everyone's doing the work end to end via agents. And no one's doing work, we're just managing agents to do the work for us. And typically you'll have five or six agents on different screens just doing the work, half an hour projects for you at a time. And that's the way things are doing.

Faye

I think I've long been an advocate of the generalists and actually that's the kind of model that you're talking about. So actually someone who can go across all of those things and can have the visibility. I think we're going to see a lot more of that moving forward.

Speaker 3

No, I was just going to reflect that. Faye, your experience is really diverse as well, right? Marketing, entrepreneurship, many different companies. Yeah, I think it really suits this kind of expert generalist. Whereas it's interesting, in the past, I feel in the workplace specialists were more favored. I'm curious if that dial is going to shift towards generalists now.

James

I find this so fascinating. It feels like in the last six months we've had a number of conversations with founders of early stage businesses. I know you're not early stage who have outwardly said the company wouldn't exist without using AI in terms of automation and just the reduction in people necessary to build the company. It feels to me, I haven't fully baked an idea on this yet, but it feels like the transformation went through compared to building a startup in the first dot com. With the advent of cloud compute and all of the kind of cost savings that brought, not having to build your own infrastructure, that was revolutionary. It feels like we're now into that next wave now where companies are being built using AI and not in traditional ways. Are you seeing that in the conversations you have with other founders?

James

Yeah, for sure.

Speaker 3

I think you're absolutely on the money there. I think we get autonomous companies. Pretty soon you might have agents just running their own companies. So there's an idea called the agent economy which is coming out now, which is the economy that is agent to agent and not really no humans are involved. The interesting question will be where do humans interface with the fully agent world, which will far outgrow the human world. Yeah. What is our role in that?

James

On the engineering side of the business, are agents writing production code for you? What kind of oversight have your engineering team got?

Speaker 3

Yeah, absolutely. Our engineers are not writing code anymore. It's fully agentic, but it's not as simple as just putting in a one word prompt. There's actually now whole processes around how to architect the agents and the sub agents so that they're accurate. There's guardrails, they reflect your business policies, your safety policies and that they're bug free. There's a whole new job role around architecting all of that. And that's the role now rather than doing the work. But that's just new work now. So.

Faye

Yeah, yeah, I feel we're teetering into a whole episode on its own here because I mean it's. Yeah, it's really fascinating stuff but I do want to bring it back to what you mentioned earlier on when we started talking about the growth of the company and that was your funding journey. So as you've said, you've been very fortunate. You've got some extremely strong backers behind you. What was that like? Was it because you'd already got a relationship with some of them or how did the funding journey actually manifest?

Speaker 3

Yeah, it's been quite interesting with us. I've never been particularly on the ball with reaching out to investors. My approach is more just build something valuable and hopefully investors will come. So I don't know if we've been lucky or not, but Cosa Ventures found us and reached out to us and then through a fairly long process, we regained their investments and then Google Ventures came off one sort of random meeting a year before. We got along well and then when it came to raising a year later, we just talked to them and it was quite a quick process. Within three weeks we had closed. Yeah, it was quite close to the end of our Runway, so it was cutting it quite fine. But that's always the way.

Speaker 3

But, yeah, my general approach is not so much to keep tabs on lots of investors, but rather just build something that's worth investing in. Yeah.

Faye

And that was the comment Christian Segestrala made at Cambridge Tech Week last year and actually the subsequent dinner. He always says, just build something brilliant and people will come knocking, rather than constantly thinking about, I've got to raise capital, I've got to think about what my exit is. So, yeah, that tallies quite nicely.

Speaker 3

Yeah. It's just the boundaries of what is good and worth investing in has completely changed as well. Now these AI companies are going from zero to one so fast that the expectations on startups have completely changed as well.

James

You're building the company in London, but you're obviously from Cambridge. So talk us through the kind of decision of how you split time. Do you have plans to put boots on the ground in Cambridge or are you going to be London moving forward? And maybe more broadly is the second part of that question. Have you had an opportunity to spend time in other kind of tech clusters and can you compare and contrast UK versus other places? Yeah, for sure.

Speaker 3

So, yeah, primarily based in London and most of our employees are. They live in satellite places near London, but we also have a good bunch in Cambridge as well and we love working from there when we can, booking office spaces there. We also sponsor Cambridge United Football Club and they also use our software. So very much involved in the ecosystem. I think Cambridge produces really well rounded individuals. I think the educational system and the culture there is really well rounded, very scientific, very technical. Where it might struggle perhaps when we compare it to say Silicon Valley, which I also spend a lot of time in because our investors are based out there, is the aggression and perhaps even the arrogance.

Speaker 3

And that's feedback I had from our investors, quite a lot actually is that I'm too nice, too polite and they want to see someone more aggressive. And at first I thought feedback was a bit silly. I think socially it is a bit silly but to be fair we have seen companies in our market grow and explode because there's some perhaps bulshy arrogant founder that is just hyping the company even if the. Even though they have no product, but the market is not smart enough to know that they have no product. So the hype becomes real. So that strategy is actually a viable strategy and it does work. That's the kind of sad part of doing business perhaps, but it's important to.

Faye

Find balance, isn't it? Between being hungry and showing that you are and being gracious. I do think it's a little bit of a British thing to be more gracious and reserved. Maybe we should start a training course on being more aggressive but I'm not entirely sure that's a good thing.

James

Yeah, again you have to you there's no question that the capital is more free flowing from the us So I guess that you have to align to some degree their expectations in the us. But I think there's definitely that mid Atlantic best of both worlds, isn't there? Where that European a little bit more measured, a little bit more thoughtful maybe?

Speaker 3

Yeah, I love talking to Europeans and doing business with them but it must be said I've been lucky enough to, to collaborate quite a lot with government in the last month or two. I was at number 10 Downing Street a couple of weeks ago talking to Keir Starmer's AI advisor and also there was a new all party parliamentary group set up just this week for innovation startups was with a number of MPs there and all the conversations that I'm hearing in government is they want to procure AI startups, they want AI startups to progress fast and it is a political issue now. But the procurement culture in the UK is just really defensive. In the US if there's value people will try you out, they'll just pay money and try it. In the UK it's a lot more around data protection just.

Speaker 3

Yeah, it's a lot more defensive.

Faye

Yeah. But at least those conversations are starting and I guess it Goes back to what you talked about earlier on. That's almost your social responsibility to make sure that you are literally knocking at the doors of number 10 Downing street to say if you want us to succeed as and stay a British born and raised company, then you actually have to invest, you have to make procurement easier.

Speaker 3

Yeah, yeah, absolutely. And we're hearing good things at the top. Was also at a press conference with David Lammy three weeks ago at the Microsoft AI conference and he kept saying that the government should be procuring AI startups should be procuring and was overly pleased to name one startup that they've piloted with in the last few years. But it doesn't flow down very well into especially local governments and all of the departments in central government are all, as we know, quite disparate and lots of bureaucracy and all the rest of it.

James

Just before we move off this kind of topic around this kind of cultural acceptance of risk and being a bit more experimental, have you seen that kind of cut through into your early customer base as well? Have you seen more adoption from US companies versus European companies?

Speaker 3

Yeah, about 35% of our client base is US, so fairly significant. And then a third is UK and the rest of it is global. They tend to be in financial hubs. A lot of Middle east actually. They're quite, yeah, quite big on big contracting projects. But yeah, US just tends to be a bit faster but also more value oriented whereas Europe and UK a little bit more relationship based. Yeah, we do get a lot of international clients because the platform is multi language, multi jurisdiction. So we're quite strong on the international cross jurisdiction contracts.

James

Am I right in guessing it's like a typical SaaS style playbook where it's self service and anyone can just sign up and get going?

Speaker 3

Yes, exactly. We do have a sales assistant motion for those who have high usage of the platform. They can upgrade by talking to us, but generally, yeah, just try out for free and then hopefully upgrade.

Faye

Brilliant. So what's next Rafi? What are you excited about happening in the next year or so?

Speaker 3

Yeah, the whole game has changed now. Every three months there's a seismic shift in capabilities because of AI agents and the product is improving so fast and we're seeing the uptick in our numbers as well, which is good. But practically speaking where we're going high personalization. So the issue with generalist Hill learned platforms is that they don't understand your company context in great detail. So really understanding the company's legal policies frameworks is one aspect. The second aspect is Building what we call deterministic negotiation place. So again, AI platforms are great for probabilistic flows. Do everything different however you want it. But in legal, you need determinism. You need rules, standards and processes. So building out these playbooks, review policies, negotiation flows on GENIE is a key priority.

Speaker 3

And the final thing is we're moving into a world where people might not even look at legal contracts in future. It's hard to imagine, but what will enable that? We believe just knowing the key issues and risks and negotiating those key points is the new interface for contracts. And so enabling humans to do that and delegating the rest to AI is the direction that we're going.

James

It's been a fascinating conversation. Just as we wrap up, what would your top piece of advice be to someone, a founder in an AI company today?

Speaker 3

I would think about workflows, not cross functionally, but linearly. So can one person do entire workflows, end to end, entire job roles, end to end? Everyone just has to think in terms of people, but in terms of agents, which, yeah, I realize there is maybe a sadness to that because what does that mean for society? But I think what it means actually is not that less people will necessarily have jobs, but you just have way more one person companies. So the game is basically how far can you stretch a one person company? That's the current game. And yeah, the way everyone should be.

Faye

Thinking, that in itself is quite an eye opener. If you think about the investment landscape, the way that we do things, that's a completely different model. So that'll give our listeners something to think about for sure. Rafi, thank you so much for joining us. It's been great having you on with us today. Where can people go and find out more about what you're doing?

James

Sure.

Speaker 3

So genie is on LinkedIn. You can search Genie AI, you can search myself, Rafi Farouk on LinkedIn or if you want to follow me personally, also on Instagram.

Faye

Perfect. Thanks for being with us, Rafi.

James

Today's show was produced by Joe Donaghy of Cambridge TV and supported by our media partner, Business Weekly. The Cambridge Tech Podcast is available on all major podcast platforms and on cambridgetechpodcast. Com. If you've enjoyed this podcast, please give it a five star review. It'll really help others discover the show.

To listen and subscribe, search for ‘Cambridge Tech Podcast’ on your favourite podcasting platform or visit cambridgetechpodcast.com.