Singapore, 11 June 2026

By six o'clock, the laptops were already open.

Before the first slide appeared, people were leaning across tables, swapping ideas, comparing notes, and asking the kind of practical questions that rarely make it into AI headlines. Over Domino's pizza and conversations that moved easily from table to table, close to 70 participants gathered for an evening that felt less like a workshop and more like a community testing what agents could actually look like in real life.

Three hours later, many of them would leave with something they could hold onto: a working first agent, a connected app, and a far clearer sense of what they could build next.

The session, Build Your Agent with ChatChat, was hosted by ChatAndBuild alongside Lorong AI, as part of Lorong AI's Tinker Time series, a hands-on format that pairs guided workshops with open build time. The premise was deliberately simple. Most people are already using AI to ask questions, write faster, or brainstorm. The next step is more personal: designing agents that understand your goals, your routines, and the way you actually move through your day.

A room full of builders

The participants arrived with very different reasons for being there.

Some wanted to understand how AI could support their businesses. Others were curious about how agents could take repetitive tasks off their plates and return more time to family and personal pursuits. One participant, a doctor, was interested in how AI might fit into healthcare workflows. At another table, the conversation centred on automating the routine in order to make more room for meaningful work.

What united the room was not a shared level of expertise, because there wasn't one. Students sat beside seasoned professionals. First-time builders shared tables with people who had been experimenting with AI for years. What they had in common was a belief that AI was not something to fear, but something worth understanding and working alongside.

They had not come to watch. They had come to build.

Setting the foundation

The evening opened with a session led by Chelsea Li, Head of Legal and Operations at ChatAndBuild, who walked the room through what ChatChat is and why it works the way it does.

She started with a distinction the company returns to often: the difference between token maxing and life maxing. The point of AI, she explained, is not to keep people glued to a platform. It is to give them back their time. That framing set up the deeper problem the product was built to solve.

Agents are powerful, but creating useful ones has remained genuinely hard. Until now, building one meant choosing between chat interfaces that are easy to use but reactive and stateless, enterprise workflow tools that are complex and admin-heavy, or developer frameworks that are powerful but demand real technical setup and orchestration. Ordinary people could talk to AI. What they could not easily do was create agents that remember, connect to tools, act, and improve over time.

Chelsea explained how ChatChat closes that gap. Beneath a deliberately simple surface sits what the team calls a serious agentic harness, the layer that absorbs the complexity so the user does not have to. She broke down what that harness actually includes: model routing for access to frontier intelligence, connectors into the tools and data that matter, persistent memory and middleware for continuity, an execution layer where real work gets done rather than merely described, proactive event streaming so agents move from answering to anticipating, and receipts and permissions so there is always a clear record of what an agent accessed and did.

The shift, she summarised, is straightforward to describe and hard to engineer. Old AI asks and answers. ChatChat remembers, watches, coordinates, and acts.

She also introduced the idea that gives the platform its name: the Living Brain. A Living Brain does not simply store information. It remembers what you chose, what you rejected, and what proved itself over time, turning isolated decisions into a system and allowing an agent to become genuinely yours rather than generic.

Learning by building

With the foundation set, the room moved from listening to building.

Chelsea and Eric led a series of live demonstrations, guiding participants step by step through the platform and making sure no one was left behind. Together the room explored how to create an agent, connect applications so the agent could pull information and take action rather than only chat, and build a working Telegram bot.

What made this stretch of the evening work was that nobody was learning alone. Members of the ChatAndBuild team moved continuously around the room, answering questions, troubleshooting, and helping people stay on track. The hour set aside for building, from roughly a quarter to seven until a quarter to eight, turned into the most collaborative part of the night, with strangers comparing screens and offering each other fixes.

By the time the sharing session began, many participants had walked away with tangible first builds, including:

A working first agent configured for a real personal or business use case
At least one connected app, so the agent could pull information or take action rather than simply respond
A Telegram bot prototype they could keep refining after the session

When participants stood up to demo what they had made, the range was striking. Different professions, different problems, different agents, all built in a single evening by people who, in many cases, had never created one before.

The story behind the mission

Christel Buchanan, Founder and CEO of ChatAndBuild, closed the evening by sharing the personal story behind why ChatChat exists.

Her father-in-law had always been an enthusiastic user of technology and AI. Over time, his eyesight deteriorated, and one day on the golf course he found himself effectively unable to see. It was then that Christel began thinking about a much larger question. What if AI could be so simple and intuitive that anyone could use it, not just the technically inclined, but anyone at all?

The goal, as she described it, is a world where someone who cannot see a screen can still access AI in what she calls this age of abundant intelligence. Using AI, she argued, should be as simple as having a conversation, in any language, without needing to read ten manuals first.

"When I met him about two months ago and took him through ChatChat, he said, oh my God, I can finally use AI properly again."

That moment, she shared, captured the vision the company is building toward: AI that feels intuitive, accessible, and genuinely for everyone. She spoke candidly about wanting to remove the barriers that so often make AI feel intimidating, and about her belief that accessibility has to remain at the heart of what the team builds. Technology, she said, should bring more people in, not shut them out.

Familiar faces, new builders

One of the most encouraging parts of the evening was how many familiar faces had returned.

Community members who first encountered ChatChat at SuperAI came back to continue the conversation. Several who had attended the Living Brain launch registered again, eager to go deeper. Students from Singapore Polytechnic, encouraged by their lecturer and drawing on earlier experiences with the community such as Codechella, joined alongside first-time attendees. Returning builders mixed naturally with people picking up an agent for the very first time.

Generations, professions, and backgrounds blended without effort. Young students sat beside experienced professionals. Builders traded ideas with curious newcomers. Strangers became collaborators over the course of a single evening. By the close, the room felt less like 70 individuals attending a workshop and more like a community learning together.

More than the technology

It would have been easy to make the evening about features. The agents, the connectors, the Telegram bots.

But the part that stood out most was the people. Their collective curiosity created a genuinely electric atmosphere, and their willingness to share ideas openly turned the room into a space for thought experiments and "what if" scenarios that played out in real time. Feedback was strongly positive, with many participants leaving energised and saying they had gained a far clearer and deeper understanding of ChatChat through building with it directly rather than hearing about it.

You did not need to be an expert to begin. You only needed to be curious enough to start.

Perhaps that was what the evening was really about. Not replacing people with AI, but empowering more people to build alongside it. The future of AI will not belong only to programmers or specialists. It will belong to anyone willing to learn, experiment, and grow with it. Judging by the conversations that carried on well past nine o'clock, and the ideas participants took home with them, that future is already taking shape.

With thanks

Our thanks to Lorong AI for hosting this edition of Tinker Time alongside us, and to every participant who came with questions, ideas, and the willingness to build something real.

ChatChat is here. To be among the first to build your own living agents, join the waitlist at chatchat.com.

Building With AI, Together: Inside "Build Your Agent with ChatChat" with Lorong AI