How Shoosmiths is turning legal expertise into AI-powered advantage
A former PA at Shoosmiths won a new car for building an AI agent that improved how the legal firm served one of its long-standing clients, showing how giving people the right tools can save time, improve consistency, and help junior lawyers learn.
Joanne Bevan (pictured) joined Shoosmiths 28 years ago as a secretary in the real estate team. Today she is a senior innovation advisor. “I’ve been at Shoosmiths longer than I’ve known my husband, 28 years this July,” she says.
She has never written code, but when the firm began piloting Microsoft 365 Copilot she was an early adopter, using it first for transcripts and drafting before building agents using Copilot’s no-code tools.
“I found it really, really easy to build agents,” she says, “but I’ve always been a bit of a techie.”
Her most recent AI agent solved a practical problem. A long-standing high-street retail client had hundreds of leased properties requiring detailed contracts to be handled with very specific procedures.
After staff changes, newer lawyers were spending too long trying to find the right documents and guidance. Bevan helped create a structured SharePoint repository, then built an agent so people could ask questions in plain English and quickly find what they needed.
“We’re not building these tools to cut jobs, we’re doing it to empower our people, help them learn, and be more innovative“
“It’s just a simple retrieval agent,” she says. “But it can’t be underestimated how powerful this can be.”
The value is not only speed but “the accuracy, knowing that we’re doing the job right and people are following the correct process” – which was also very reassuring for the client, she says.
The idea won Shoosmiths’ internal client excellence competition. But more important than the prize – a brand new car – was what it illustrated: anyone in a business can come up with useful ideas, and you don’t need to be an engineer to solve problems using AI.
Gold standard agent
David Jackson, Shoosmiths CEO, says the firm has “gone all in” with Microsoft technology, giving people access to Microsoft 365 Copilot and Copilot Studio and encouraging them to find new uses for it.
Contract review is one of the clearest examples.
Tony Randle, director of client technology and service improvement, says Shoosmiths lawyers review hundreds of contracts every year. Before generative AI, the firm had already built a natural language processing system for commercial supply agreements, but “it took us two years … to train that system” on a single contract type, he says.
Large language models offered a different route. But instead of teaching an external tool to think like a lawyer, Shoosmiths decided to codify its own expertise. The result was Apollo, a contract review system that applied Shoosmiths “playbooks” to draft contracts.
Randle describes these playbooks – which can contain hundreds of rules – as “the aggregate of scores of lawyers’ experience … distilled down to a gold standard”. They set out what a contract should contain, why it matters, and how to fix problems using tried and tested drafting.
Built on Microsoft Azure, Apollo uses AI models to compare draft contracts with those playbooks. Randle says the system is “brilliant” at reading the rules and the contract and spotting where a rule has not been followed, but “it doesn’t need to be a lawyer”.
That mattered because the firm could not find an off-the-shelf product that met its standards. After reviewing around 15 suppliers, Shoosmiths decided to build its own and Microsoft introduced the firm to prospective development partners.
Transparency and learning
A key feature of Apollo is its transparency. Rather than operating as a black box, the system shows lawyers why it is flagging an issue and which part of Shoosmiths’ own playbook it is drawing on. For a profession built on judgement and trust, that transparency matters as much as speed.
While Apollo is still in pilot phase, feedback has been positive so far. Sarah Hartley, a trainee solicitor at the firm, says: “I found Apollo really valuable as a learning tool – it doesn’t just assist with marking up a lease, it identifies issues that I might not otherwise have spotted, suggesting practical drafting solutions and providing a clear explanation of why it has flagged those provisions for change.”
This transparency is an essential element of Shoosmiths’ AI philosophy, says Jackson.
“We’re not building these tools to cut jobs, we’re doing it to empower our people, help them learn, and be more innovative. And it means we offer our clients something they can’t get elsewhere,” he says.
Diffusing knowledge, driving efficiency
There is also a strong efficiency case. Drawing on earlier AI work, Randle says this kind of review can save three to five hours on a single legal document – a significant saving when you understand that the team piloting Apollo handles about 3,000 documents a year. During periods of heightened mergers and acquisitions activity, Shoosmiths has advised on more than 500 corporate transactions in a single year, says Randle.
The firm is now extending the approach to more document types and teams. Randle says 63% of partners who responded to an internal poll said they wanted to be early adopters.
For Bevan, the lesson is simpler: give people the right tools and you’ll be amazed by what they can achieve.
That is what makes the Shoosmiths story distinctive. It is a firm using AI to capture and diffuse knowledge, empower its people to learn and innovate, and give it a distinctive edge in a competitive legal marketplace.