“It didn’t feel foreign anymore”: Kim Teo on how me&u became an AI

Restaurant technology startup Me&u has unveiled its new reservations platform, a massive product update that CEO Kim Teo says was accelerated by the use of AI.

Its launch coincides with a surge in ‘AI-first’ startups, claiming to have fundamentally reshaped their business around emerging digital tools.

This month, Neural Notes spoke with Teo to hear more about the company’s internal transformation, what AI means for traditional team structures, and what others can learn from Me&u’s own ‘AI-first’ update.

Her take on AI-powered product development was clear.

“If you wait around for someone else to prove it out, you’re probably already a year late,” she said.

The startup’s reservations platform adds a new dimension to the business, perhaps best known for its QR code table-ordering technology and its guest data platform.

It comprises three functions, the first being its core reservation management platform, challenging global players like Resy, SevenRooms, and OpenTable.

On the customer side, Me&u’s reservations platform will allow guests to book a table at their favourite venue.

Restaurants will see those bookings mapped across their own floor plan, helping them maximise revenue and handle table turnarounds efficiently.

The second key function is an AI voice assistant, capable of handling most inbound booking requests without a human staff member on the line.

With its new platform, Me&u is also branching out into function management, giving guests and operators a clearer view of what is booked, for when, and whether payment is required.

Teo said the three functions can draw from the same pool of guest information, providing a more seamless experience for frequent diners — and the venues struggling to connect various booking, CRM, and management systems.

“Running restaurants has gotten to a level of complexity where you’ve got five, six, seven different platforms that all touch your guest experience,” said Teo.

“The job has become cobbling together an integrated stack of systems.

“That is fine for a startup tech company like us, but it’s not fine for a restaurant operator, where they would prefer for it to be all-in-one rather than separate multi-products.

“So that’s the suite of products that we just launched.”

AI capabilities change what is possible

Customers will find artificial intelligence capabilities woven through the new tools, not simply ‘bolted on’ to legacy systems, according to Teo.

The Me&u team also harnessed AI when building the platform, she added.

It is de rigeur for founders to claim ‘AI-first’ status, with founders hoping to show customers and investors they are effectively utilising the latest generative tools.

But Teo says Me&u really did reshape itself around AI tools, starting in August 2025, when it tapped a small team of engineers for a new, stealth project.

“We had three people who were really keen to build something, and credit to them, they were hustling me to build something transformative for the business,” said Teo.

“They didn’t want to build a little module, or a widget, or work on automation workflows for customer service, which we have done.

“They wanted to build a product that could be as big a business as our core ordering and payments product, or even bigger.”

That small kernel of “AI optimists” worked with the help of AI coding assistants, which provided support equivalent to a “really good junior who wouldn’t stop working”.

At that point, the plan was to build the standalone reservation platform, without venturing into voice AI and events.

That trajectory changed when Me&u informed some select restaurant partners of the plan, and their input informed the platform’s expansion.

However, the most significant change came in late 2025, with the public release of next-generation AI tools like OpenAI‘s GPT‑5.2‑Codex.

“Past December and January, it started to become, ‘Oh no, they’re way more than a junior,’” said Teo of the AI coding assistants.

“There was a significant change in capability from a model perspective.”

That dramatic capability uptick correlated with increasing token usage.

Me&u AI engineer Ryan Hendler personally rocketed to #12 on the global list of ChatGPT Codex’s top token users, according to the Australian Financial Review.

Taking the team along for the ride

Allowing a small team of self-confessed AI advocates to use new tools is one thing, but convincing an entire startup to adopt a new way of working is another thing entirely.

“There’s initially a lot of ‘This seems cool, but what does it actually mean for my day-to-day? How does it change the way I work?’” said Teo.

To bring the reservation team’s findings to the broader business, Me&u leadership in April asked its workforce to pause all non-essential engineering and product design to experiment with emerging AI tools.

“It was a full month dedicated to learning, upskilling, and thinking about how things would change the process,” she said.

“We came out of that with, I think, a far more engaged and excited team. It didn’t feel foreign anymore.”

Everyone has a point where they “see the light,” claimed Teo. What varies is how far leadership must walk with them beforehand.

“It was definitely a cultural shift, and it’s taken a bit of time,” she continued.

“But having a project like reservations in-house has also been a great forcing function.

“Like, ‘Hey, this isn’t just what you’re reading on X or Reddit. This is something that’s happening internally, and you’ve got team members who used to work in the old way that have now transitioned to working in the new way.”

Does AI-first mean people-last?

Beyond questions about how people work, surging AI capabilities raise questions about whether people will work at all, or if their tasks will eventually be subsumed by highly capable digital tools.

The critical view is that by implementing advanced AI systems, businesses are tasking workers with using and training tools that could eventually replace their roles.

For Teo, who has presided over Me&u and predecessor Mr Yum through redundancy rounds in 2022 and 2023, the promise of AI is not in reducing headcount.

It’s not about “replacing humans”, she said, “because every company’s backlog is massive, and all the things that you want to do you can never do”.

However, Teo believes the days of engineering teams with hundreds of staff are no longer necessary for ambitious, global software companies to make their mark.

Smaller engineering teams are often closer to customers and feel more accountable to them, driving better results, according to the CEO.

That equation changes somewhat for go-to-market teams, she added.

“Especially in hospitality, human-to-human sales is going to be around for as long as I can see,” said Teo.

“There’s so much shit in your inbox now that no-one’s replying to… when someone hasn’t taken the time to write it, you don’t feel like you need to take the time to reply.”

The company has since doubled down on Teo’s view that there are some roles AI cannot replace.

Just days after Teo spoke to Neural Notes, Me&u announced Tobin Hartley, formerly of SevenRooms, as its new director of ANZ sales.

Becoming an ‘AI-first’ startup doesn’t mean shutting the door to younger staffers with less experience, either, said Teo.

The CEO reflected on three qualities — hunger, humility, and smarts — that matter in new hires, above and beyond their ability to churn out code as effectively as an AI copilot.

“Things will continue to look very good for juniors [who] are ultimately creative, curious, wanting to be customer-facing, hungry, work hard — all of the things that have always worked,” she said.

To prove the point, Teo said a 21-year-old Me&u team member was among those to bring its reservation platform to market.

All of that AI use comes at a cost.

As the appetite for compute grows, startups are spending more than ever on enterprise-grade services from the likes of OpenAI and Anthropic.

Open-source models do promise alternatives to subscriptions and API access deals with the big-name providers, but for now, more AI usage generally means higher costs.

For Teo, that spend is justified.

She said that using AI effectively can triple or quadruple the productive output of a mid-range engineer.

“Actually, when you look at the spend, it all feels like it makes a lot of sense,” she said.

That doesn’t mean unchecked usage is appropriate.

The era of “token-maxxing”, where engineers effectively competed to see how much compute they could use, has ended, said Teo.

Despite its hunger for AI access, the Me&u business keeps its costs “relatively controlled” by opting for subscription models over full-blown API access, she added.

Neural Notes is a weekly AI column that looks at how AI is influencing Australia.

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