Patent protection has always carried a structural disadvantage for fast-moving companies. Filing requires stopping, engaging an external attorney, and entering a slow and expensive process. Dominic Davies, founder of Lightbringer, started working on this problem in 2013, long before large language models made it tractable.

Dominic had an unusual profile for the task: he trained as a software engineer before qualifying as a patent attorney, then spent years building automation tools in his spare time. Natural language processing offered early promise but could not clear certain key hurdles. That changed in 2022 when he gained access to GPT-3. "I could see what it could do, and it solved several of the problems I had been stuck on for several years," he explains. What followed was both excitement and anxiety: it was now clear that significant parts of legal work would eventually be automated.

Lightbringer's first target was patent law firms. The team approached roughly ten firms, and the response was consistently negative. Firms could not see how to manage a service that was now faster and cheaper: if a task took half the time, they would earn half the fee. The model did not fit. The pivot came from technology companies themselves. These companies were already running their businesses on software platforms and were looking for an AI-native approach to intellectual property management. Nothing existed. Lightbringer built it for them.

The structural difference between Lightbringer and a traditional patent firm comes down to incentives. A Sequoia Capital investor captured it precisely: "the future of legal is software companies masquerading as law firms." Lightbringer is built by software engineers, priced on subscription and flat fees, and designed to deliver a faster and better outcome at each iteration. The billable hour, which rewards time spent rather than value delivered, has no place in that model.

The platform supports the full patent lifecycle: drafting, prosecution, portfolio management, and strategy. Users can sign up for a free account, import existing patents and inventions, and work through a guided process at their own pace. For more AI-forward users, Lightbringer supports MCP (Model Context Protocol) integrations with Claude, ChatGPT, Jira, Linear, and Slack, allowing engineers to delegate patent management to their own AI agents directly. Those agents can read patent office correspondence, pull in engineering data from connected tools, build responses, and take strategic decisions about filing order and prosecution approach.

Underlying the service is an agentic drafting workflow with at least 46 steps: research, drafting, review, checking, and inventor discussions. The agent handles all of it; certified attorneys review the output, check scope, and provide the legal sign-off required by law. Davies calls this the auditor-in-the-loop model, and describes it as a mixture of transitional and permanent design: some legal tasks will eventually be fully automated, while others will retain attorney involvement as long as legal liability and professional standards require it.

On the competitive landscape, Davies draws a clear line between AI-native companies and those building on generalized models. OpenAI, Anthropic, and Google will continue building general-purpose tools and will never focus on the specific workflows of patent prosecution. The companies that understand their customers deeply enough to build those workflows will hold a durable advantage. Law firms, meanwhile, are splitting into three groups: those fully committed to change, those experimenting cautiously, and those with their heads completely in the sand. Davies expects the third group to disappear.

"Today, Lightbringer software does 99% of the work that some patent attorneys still do at other firms," he says. Five years from now, he believes the industry will be unrecognizable. Lightbringer, built with what he describes as "a Silicon Valley mindset," intends to be the market leader that shows the rest of the industry what the new standard looks like.

For more on how AI is reshaping decision-making in technology investments, explore the Dedale Intelligence article library and the September 2026 Market Intelligence Report.

Contact Dedale Intelligence to learn more about our research coverage.

This interview was conducted by Miguel Tang. Connect with Dominic Davies on LinkedIn to follow Lightbringer's progress.