Create a free account, or log in

Use your email or the options below

By continuing, you agree to our Terms & Conditions and Privacy Policy.

Or

Want unlimited access?

Get your intro offer. 

Relevance AI bet big on AI agents before big tech caught up

AI Agents have been the talk of the tech world over the past few months. But this Australian startup was ahead of the curve over a year ago.
relevance ai
L-R: Relevance AI co-founders Daniel Vassilev, Daniel Palmer, Jacky Koh. Source: Supplied

A year ago Relevance AI secured $15 million in a Series A funding round. While most of the tech world was still plugging generative AI, the Aussie startup was looking beyond copilots.

At the time, the term ‘AI agents’ was not yet on the lips of every major industry player. But Relevance AI was already betting on it.

While not a new concept, it wasn’t until September 2024 that the term ‘AI agent’ became more ingrained in the mainstream. That week saw a perfect storm of Hubspot and Salesforce conferences where both SaaS companies announced their AI agent offerings – with Google also getting on the agent train simultaneously.

To be fair, AWS did announce AI Agents last year as well. But otherwise, the back half of 2024 has been the real tipping point for this evolution in generative AI discourse.

“It does feel good to see the industry adopting it. Because from our perspective, it legitimises us in the business. Buyers are much more interested,” Daniel Vassilev, co-founder of Relevance AI said to SmartCompany in Las Vegas last week. 

Relevance AI participated in AWS’ Generative AI Accelerator this year, which Vassilev described as a chance to “work more closely with the team” and explore potential collaborations.

According to Vassilev, this head start wasn’t just luck, but a move grounded in years of experience. 

Vassilev and his co-founders brought a deep understanding of automation, having built tools that handled complex workflows long before agents became a buzzword. 

Relevance AI’s vision for an AI workforce

For Relevance AI, it is this understanding that is crucial for businesses to be successful with AI agent automation. 

Relevance AI’s approach to AI is teams of agents designed to handle workflows autonomously, rather than just assisting with individual tasks. This philosophy stems from its belief that businesses should be constrained by ideas, not headcount.

“A lot of people building in this space are focused on engineering and frameworks. We’ve departed from that because a lot of people don’t realise that automation doesn’t usually fail because of technical limitations. It typically fails because of what we call a lack of organisational wisdom,” Vassilev said.

Vassilev said if a company has processes that aren’t properly understood outside of subject matter experts – and aren’t captured anywhere – then trying to automate them isn’t going to work.

“We felt the more that the AI workforce can absorb that hidden knowledge, the better and more capable it becomes,” he said.

This focus on empowering subject matter experts is reflected in Relevance AI’s low-code tools, which allow non-technical teams to create custom agents tailored to their workflows. The platform also addresses key enterprise concerns like security and reliability, with features like SOC 2 Type II compliance and permissioning that mirrors employee access levels.

Since launching its agents product last year, Relevance AI has seen adoption across industries, with use cases ranging from sales and recruitment to operations. 

According to Vassilev, recruiters often spend significant time manually checking CVs for specific information – a process he says AI agents can streamline by extracting relevant data and populating applicant tracking systems.

Vassilev says this automation allows recruitment teams to focus on higher-value activities, like building relationships with candidates, while also enabling them to process more applicants

The co-founder also used sales as an example, where an AI agent can do hours of account research in a fraction of the time while humans can focus on other tasks and relationship management.

When it comes to the notion that AI is going to put people out of work, Vassilev disagrees.

“I’m extremely convinced that whenever you give people the opportunity to do more things, they don’t decide ‘I’m going stop working on something… they typically just do more,’” he said.

“History has shown that we’re pretty good at maybe being a bit greedy as humans.”

While Relevance AI may have been early to the AI agent train, it now has more competitors coming in from the big end of town. 

But Vassilev remains confident in the face of big tech.

“The thing hopefully you’ll be asking us a year from now is ‘how were we able to keep up with that and establish ourselves the market leader?’,” Vassilev said.

The author travelled to re:Invent in Las Vegas as a guest of AWS.

Never miss a story: sign up to SmartCompany’s free daily newsletter and find our best stories on LinkedIn.