Sooner or later, almost every big company builds a customer help centre.
They fill it with how-to articles, sort them into neat little categories, add a search bar, and hope customers can find what they need. Almost nobody reads it.
Daniel Kimber had lived this from both sides: first in support, where he watched users hit the same walls again and again, then again in product, where he watched teams pour hours into forgotten documentation.
When he and co-founder Ajain Vivek started asking what AI could do that help centres could not, the answer became Brainfish – a platform that now resolves support problems for millions of users each year
Daniel found a side door into product
Daniel grew up playing video games and, by high school, had started pulling them apart in his head, wondering how they had been made.
He taught himself to code, built computers at home, and with a couple of mates created a digital awards system for his school.
“We won the UNSW Innovation Award for that,” Daniel says, “in, like, Year 10.”
At eighteen, he landed a part-time support role at SiteMinder (a hotel software startup that’s now an Australian unicorn) and became one of the youngest people in the company.
At the same time, he was studying information systems at university, partly because the Atlassian founders had done the same degree.
"I thought, that's a good starting point for me," he says.
He wanted to work in product, but no one was going to lift him straight from the support desk into a PM role.
So he pitched himself as a software engineer – "I got a reality check quite quickly," he admits.
But his future co-founder Ajain Vivek, then a tech lead, was looking for a Quality Assurance Engineer and saw potential.
Daniel spent the next four months volunteering one day a week with the team until SiteMinder finally offered him a formal Quality Assurance role. Then, in his final year of uni, he made another push and talked his way into a junior product position.
“It took me about three years,” he says, “but eventually I got in the door.”
Brainfish came from a familiar pattern
After several years working in product across different companies, a recurring pattern had become impossible to ignore.
“The best customers, the top ten per cent, were the ones really using the product properly,” he says. "And they were making thirty per cent more revenue than everyone else."
When he looked closer, one thing jumped out: those customers had all been formally certified with a course to learn how the entire product worked. The other ninety per cent had not, and there was no other obvious way to train all of them.
"Software gets really complex really quickly, especially in B2B," Daniel says. "And often the people using it aren't technical."
The usual fixes, like help centres and FAQ pages, still left plenty of users stranded. There was always a gap between how much a company knew about its own product, and how much a user could find or understand. Daniel and Ajain started wondering whether AI could close it.
The start of Brainfish
On the side, the two of them had been experimenting with GPT models and reading the research papers as they dropped.
They had also just spent about a year building a completely different startup, an employee benefits company that, in Daniel’s words, failed in almost every way possible.
"We built the solution before we had any customers," he says. "We built for a problem that we thought existed, but didn't really."
The failure stung, but it taught them exactly what not to do again. So when the Brainfish idea surfaced, they did the opposite, they pushed hard on the problem first before they built out a full product.
They believed that if AI could understand the product well enough, the features, the workflows, the weird edge cases, it could answer any questions a user would likely ask.
“The idea was to take all this product knowledge and build an instant answer experience,” Daniel says, “to make it easy for the average person to understand a complex product.”
Then, only weeks after they started prototyping, ChatGPT launched.
“We had literally been pitching a slide deck that showed: ask a question, get an answer,” he says. "And then suddenly there was a consumer version blowing up in everyone's face."
Instead of killing the idea, ChatGPT poured petrol on it: "It helped us to just really push for this solution," Daniel says.
The hunt for the first customer
Daniel spent two to three months working through his network before he found Mad Paws, an ASX-listed pet services marketplace with a flood of customer questions and enough curiosity to experiment.
The opening came through Mad Paws’ Head of Product, someone Daniel knew from his SiteMinder days.
“He said, I’m not actually the right person, but I know the Head of Customer Experience, and he’s a good friend of mine, so why don’t you talk to him?” Daniel recalls.
That one introduction turned into a pilot for Branfish.
"I think networks are overrated and underrated at the same time," Daniel says. "Having executive-level networks is very hard, but it pays dividends the more you go through the startup experience."
The results from that first pilot were significant. Within four months, Mad Paws had run more than 30,000 customer questions through Brainfish, cut inbound support tickets by 22 per cent, and recorded a 636 per cent return on investment.
One of the biggest wins has been the team
Three years in, Brainfish has raised $15 million across three rounds (including a Pre-Series A led by Prosus Ventures), expanded into the US, and built a customer base across six countries.
When Daniel moved to San Francisco twelve months ago, the US made up ten per cent of revenue. Now it’s well over half.
But when Daniel talks about what his greatest success is, he talks about the team.
“Our team is probably my biggest win,” he says. “We’ve got a team of 25 people now, and the fact that I would back every single one of them in their role a hundred per cent, I think that’s fantastic.”
The early engineering hires came more easily because many were people Daniel already knew, including former SiteMinder colleagues. But once Brainfish had to hire commercial leaders and customer success people through open searches, the margin for error got much thinner.
With help from one of their investors, Brainfish built a more disciplined hiring framework.
"One of the best things to do with most commercial hires is to get them to do a presentation," Daniel says. "You can see their ability to communicate, their thought process, their problem-solving ability."
It mirrors the way Daniel earned his own first chances: do the work first, then let the title catch up later.
The pressure never stops
"One of the challenges is that you never switch off," Daniel says. "Everybody is always thinking about work all the time. No one really quite understands that unless they've also been a founder."
When you work for someone else, even in a hard job, there is still some boundary to it: weekends, leave, holidays. You can take a break: "But when you're running your own business, you cannot," he says.
Daniel’s answer has been routine: exercise, time blocked out to decompress, social plans he protects, and friendships outside tech so the whole week does not shrink into one long conversation about work.
"I think exercise is a really good way to clear the mind," he remarks.
The other long-running challenge has been hiring, moving quickly without bringing in the wrong person. "It's easy in the early days when you are desperate for a role to make a decision quickly," Daniel says.
Often the problem was not the candidate. It was that the role itself was still blurry around the edges.
"With one of our roles, we didn't know the full scope of what that role needed to be,” says Daniel “It took us probably a few months to really scope that out well before we were able to recruit properly for it."
His hard lesson was to slow the process down, get brutally clear on what was needed, and only then start searching.
"We really knew what we were looking for," he says, "and that let us write good requirements and filter properly."
Stone & Chalk helped in the messy early stage
In the first year, when almost every part of the company still felt loose and uncertain, Brainfish worked out of Stone & Chalk in Sydney.
For Daniel, the best value was being around other founders who made the whole process feel less lonely.
"Coming to a place where everyone is working on the same type of problems is really useful day-to-day," he says. "Having other people around you that are all on that same kind of stage makes a massive difference."
Stone & Chalk also helped with the hard everyday questions that founders often don’t see coming until they are knee-deep in them.
“Stone & Chalk was really helpful with some of the early things we were trying to figure out around insurance and logistics and setting up some of the core business stuff that’s painful to work out when you haven’t done it before.”
The advice he would give founders now
“You can build a business much bigger than you think with a small team,” Daniel says. “But you also don’t have to kill yourself over building your company, you can make it a sustainable adventure.”
Startup culture often rewards exhaustion as proof of commitment. Daniel does not reject the intensity, he is clear about how consuming it really is, but he draws a line between intensity and self-destruction.
"Life is long, and you don't want to ruin it over a passion. It's supposed to be something you want to do."
He also talks about how much easier the work becomes once the fog starts to lift.
"When you’re starting the company, you have no idea about anything, so you’re just trying to figure stuff out. By year two, reality hits. By year three, ideally you’ve figured out exactly what you need to figure out.”
"I'm really excited because we know exactly what problems we're trying to solve and we're mostly just focused on execution."
From support desk to startup
Brainfish started with a pattern Daniel couldn't ignore: companies were building powerful software that most customers barely used.
Closing that gap meant drawing on everything his career had taught him – that the best way in is to prove yourself first, then let the title and opportunities catch up later.
It's a long way from the support desk where he started. But the problem underneath it hasn't changed: someone is stuck, the answer is somewhere in the system. Daniel just wants to help them find it.
