Unicorn patent intelligence platform PatSnap harnesses GenAI and customer needs for growth

*This article is adapted from * The Business Times
Access the print version here.

[SINGAPORE] As reports circulated on how artificial intelligence could kill off white-collar jobs, PatSnap co-founder and CEO Jeffrey Tiong turned anxious.

It was only after learning and understanding more about AI that Tiong’s anxieties about it potentially allowing someone else to create a competitor to Patsnap’s patent intelligence platform abated.

PatSnap’s patent intelligence platform helps companies to avoid infringing existing patents with their new products while being able to see the direction in which competitors are developing their patents.

However, the company’s customers are enterprises, which have a range of considerations when they use the software.

“They don’t just want a cool user interface, they want data security (and) undisrupted existing workflows,” Tiong told The Business Times.

The platform became a unicorn in 2021 after a funding round that included Sequoia China, Shunwei Capital and Vertex Ventures. From a Singapore startup, it has expanded across 50 countries and in key R&D centres such as Shanghai, London and Toronto.

Competitive edge

Aside from the needs of enterprise customers, PatSnap’s database gives it a competitive edge. The company has a clean database of patent information that can be used to train AI models. The patent intelligence platform is model agnostic, allowing customers to choose whichever AI model they want.

“We have curated a high quality unique data set that differentiates us from the generic AI models,” said Tiong.

Aside from training AI models, PatSnap has built AI agents to specifically help with issues such as patent drafting. Tiong recounted an example where the AI agent PatSnap created aided a life sciences customer to better identify a drug target and its effects.

The advent of generative AI (GenAI) has allowed PatSnap to cater to a new market of R&D departments after previously serving only patent departments. R&D engineers can query the PatSnap database using GenAI, which will also make it easier to understand the results without the patent jargon and terminology.

This new business came about after feedback from its patent department customers, who were seeking to work more closely with the R&D teams. Patent departments wanted the R&D teams to think about intellectual property (IP) at the start of product development rather than during the process or after the product was finalised.

Operating across industries and countries, PatSnap is able to help R&D teams come up with ideas that may not originate from their industry. Tiong cited the Dyson hair dryer as a reported example of an idea going across industries, with its genesis from an industrial fan.

“Because we collected all the patents from around the world, across industries, we can recommend innovative solutions to R&D engineers and help to open up a much bigger market,” he added.

Working with the R&D department means that rather than just serving the typically one person in the patent department, PatSnap can also engage with a company’s few hundred-strong R&D teams.

There are no plans to go public anytime soon, according to Tiong.

However, with funding taps drying out post-Covid, he told employees to hunker down and focus on growth, and PatSnap has since delivered.

The patent intelligence platform has been able to keep growing, with a 20 per cent average growth rate in revenue since 2023, crossing US$100 million in annually recurring revenue in 2024. He also said that PatSnap has also delivered on the profitability front.

“The last few years we have been doing well on both fronts, and if we can do well on both fronts, be it private or public funding, we are good,” said Tiong.

To the top

PatSnap’s embrace of AI stems from the top, as Tiong believes that someone who uses AI will be better off than someone who doesn’t. He has used AI to bounce ideas and vibe code simple apps for his own use. Vibe coding is a hands-off approach to software development, by using plain English to prompt an AI to generate code.

“This type of AI transformation has to start from the founder and CEO. I myself have to learn to use it,” he said.

At PatSnap, presentations are no longer done in PowerPoint decks. Instead, AI is used to generate HTML code for discussions. This process of changing the mindset of people within PatSnap has not been easy, with the need for both carrot-and-stick approaches, added Tiong.

Employees are given about US$1,000 in tokens to use a month, with Tiong encouraging “tokenmaxing” to the cap. There is also a monthly incentive, with winners picked from teams, individuals and leaders for the best and most innovative AI application.

“It is really telling them that we have to change if not we will get disrupted and obsolete,” said Tiong.

PatSnap is navigating a period of opportunity and risk, needing to catch this wave of AI to ride on, he added. Equipping PatSnap’s employees to enable them to use AI will be key, as companies such as PatSnap now target not just IT budgets but also service budgets at enterprises.

In other words, companies that buy AI agents from PatSnap are no longer tapping IT budgets but budgets that would have been used to engage third-party service providers such as consultancies.

Now Tiong wants the company to be the No 1 patent intelligence platform in the world, and surpass anyone who is trying to use AI to vibe code an alternative.

“PatSnap being founded here and grown globally... we are one of the few rare Singapore companies that go global and can be No 1 in our space in IP and R&D intelligence,” explained Tiong.

For the latest news on Vertex Ventures SE Asia and India and our portfolio companies, follow us on Linkedin or subscribe to our monthly newsletter.