Keith raised £2 million to launch AI law firm to accelerate property deals by 70%

British legaltech startup Keith has attracted £2 million in seed investment and plans to launch a regulated AI law firm by the third quarter of 2026. The initial focus is handling residential property transactions, where the company promises to automate the bulk of routine work and reduce deal closing times by 70%.

Where Keith is entering the market

Conveyancing is the legal work involved in buying and selling residential property in the UK, including document verification, correspondence with parties, registration of property ownership transfer, and settlement with government bodies. For the local market, this is a pain point: according to the company's estimate, over 530,000 property transactions fail annually, and one reason is the slow, opaque, and labour-heavy process. The startup was founded by Andy Shovel and Pete Sharman, who previously built the food brand THIS, with Sam Tucker as the third co-founder responsible for the product.

The seed round was led by Backed VC, with participation from Breega and angel investors. Keith intends to first obtain regulation through the Council for Licensed Conveyancers — the UK's professional body for property transaction specialists — and later, when expanding into other practice areas, pursue a broader regulatory status.

How the model works

Within Keith, a network of specialized AI agents is being built to handle document verification, draft preparation, client communication, and deal stage management. The company claims that such a model can automate up to 80% of the work previously done by humans, but critical decision points remain with qualified specialists. The idea is not to remove lawyers from the process, but to keep them in control of disputes, judgement calls, and risk management.

• Verification of title documents and incoming correspondence
• Analysis of property search results and other checks
• Preparation of responses and document drafts
• Real-time deal status tracking
• Client communication 24/7 via phone, WhatsApp, and other channels

"The client-facing agent will be almost indistinguishable from a human," is how Keith describes its service layer.

Special emphasis is placed on customer experience. Instead of long email chains, hold times, and infrequent updates, the company promises constant availability, instant responses, and a clear next step for each party in the transaction. For a market where clients often don't understand what's happening between submitting documents and registration completion, this may be as important an advantage as back-office automation itself.

Why this matters

Keith's story shows that legaltech is increasingly moving from "copilots for lawyers" to full AI-first firms with their own license and operating model. It's no longer just about speeding up the preparation of a single document, but about rebuilding the entire service: from initial client contact through data exchange with registries and tax systems via APIs. For the founders, this is a particularly logical entry point: conveyancing is mass-market, standardizable, and comparatively formalized.

Keith also has an ambitious quantitative target: reducing deal time by 70%. If the company even approaches this figure, it will be a strong argument for the new model. The key question, though, will not be the quality of the demo, but how robustly such a system handles regulatory requirements, edge cases, disputed scenarios, and human expectations.

Legal services are unforgiving of errors, so the combination of automation with mandatory human oversight is not a marketing detail here, but a fundamental condition for survival.

What this means

Keith is betting on one of the most conservative segments of the legal market and trying to turn it into an AI-managed assembly line under the supervision of licensed specialists. If the launch in the third quarter of 2026 succeeds, the market will receive yet another strong signal: AI in law is no longer sold as office software, but as a full-service offering with new demands for speed, transparency, and accessibility.