Meet
A physical community for people who want to compare notes, find collaborators and make AI less remote.
A Melbourne AI community / working page
A place to
build with AI.
The Melbourne AI Hub is developing a practical, welcoming community where people can meet, learn, demonstrate, discuss and build with artificial intelligence.
Level 3 · 480 Collins Street · MelbourneExplore the Hub
Founder, People and Robots Pty Ltd
A personal note
Proud to be
in the room.
“I’m proud to be a founder of the Melbourne AI Hub. This is where the ideas I’m working on can meet other people, become practical projects, and stay connected to the city we’re trying to serve.”
01 / The proposition
AI is arriving
in the room.
There is plenty of AI news online. The Hub is interested in what happens when people sit together and try things: a demonstration that answers a real question, a business problem made less abstract, a robot considered in the context of care, or a hard conversation about what these systems are doing to ordinary life.
It is for experienced practitioners, curious newcomers, businesses, students, researchers, creatives and anyone who wants to take part in Melbourne’s practical AI community. The Hub is still taking shape, which means the people who show up can help shape what it becomes.
Make
Practical demonstrations, experiments, businesses and prototypes—not just predictions about the future.
Discuss
Serious conversation about society, work, care, robotics and the choices that should remain human.
02 / Why I’m involved
A community
for the work
in between.
I’m Martin Davies-Roundhill, founder of People and Robots Pty Ltd. I was invited to join the Hub’s founding work because the questions I’m already pursuing belong in a room with other people: how intelligent systems can support care, how agents can work together, and how we keep humans accountable while machines become more capable.
The Hub gives those questions a local setting. It connects my work on robots, agents and public-facing experiments with people who may approach the same transition from business, research, creative practice, government or everyday life.
Involved in the founding work · public details still developing
03 / Projects I’m bringing to the table
Make the
systems useful.
These are working projects and proposals, not promises that a finished service is already running.
Meeting assistant · prototype direction
A bot that can sit in the meeting
The first practical version would help prepare the agenda, join the meeting context, post timing and agenda prompts where supported, and turn a visible Teams transcript into decisions, unresolved questions, owners, due dates and minutes.
A human chair stays in control. Any recording or transcription should be visible to participants, and minutes or external follow-up should require human approval. Spoken participation and several equal conversational agents are later possibilities, not current claims.
Bot-to-bot communication · repair and experiment
Agents that can talk to one another
I’ve been working on a shared message bus so bots such as Nicabot and Scientist A can send tasks, acknowledgements, results and failures to one another with a recoverable record of what happened.
Telegram is currently part of the workaround and the human-visible audit trail; private bus delivery and a Telegram post are separate things. The next useful step is a harmless end-to-end diagnostic, followed by a reliable bridge into the meeting assistant.
People and Robots · connected work
Embodied intelligence in service of care
People and Robots follows humanoid and social robotics with a practical question: what would actually help people, families and carers? The Melbourne AI Hub is a place to bring those questions out of the product demo and into shared discussion.
Explore People + Robots04 / How the work should feel
Useful.
Visible.
Human-led.
AI participation should make a meeting clearer, not make people wonder what was recorded or who made a decision. That means visible systems, recoverable messages, explicit permissions and a human chair who can stop or correct the process.
- Do not silently record or persist meeting media.
- Do not invent decisions, owners or commitments.
- Keep Telegram publication separate from private bot transport.
- Show evidence when a system has actually delivered something.
Join the conversation
Bring a question.
Bring a demo.
The Hub is developing at 480 Collins Street, Melbourne. If you’re interested in practical AI, agents, robotics, business, care or the social consequences of these systems, get in touch.
Send a noteThis is Martin’s working page about the Hub. Details such as membership, governance, events and public channels may change as the community develops.