What Apple did to PCs, we're doing to AI
In the late 90s, Apple took a deeply technical category, personal computers, something that was seen as nerdy, hobbyist, and inaccessible, and made it intuitive, desirable, and mass-market.
They did it by obsessing over product quality and by hiding the complexity from the user.
We believe AI is at a similar moment right now.
Earlier this year, OpenClaw took the world by storm.
Like every other curious engineer, my co-founder Ricardo and I jumped in to see what we could build with this new technology - the results were incredible.
We both built things that were genuinely useful in our own lives.
And every time we showed them to friends and family, we kept hearing the same thing: how do I get this thing!?
Very quickly, that turned into a whole set of questions.
How do I get my own VPC? Should I buy a Mac Mini? How much is this going to cost? What subscriptions do I need? Is it secure? What are all these issues people are talking about on X? How do I even set it up?
It became obvious that this was something people really wanted. The virality of the product made that clear to everyone.
It also showed something even more important: people were willing to trust these agents, because they could actually do real work.
But the effort required to get everything running, and then keep it running, was completely out of reach for a regular person.
That was the moment we started asking ourselves: how do we get this technology into the hands of everyday people?
How do I get my auntie, my mum, my sister to use this technology? What does the world look like when every person has their own personal AI?
After meeting six months ago and realising we were both independently thinking about, and building toward, the same problem, we decided to tackle it together.
Since then, we've worked with hundreds of people, helping them understand their AI needs, while also building and using these systems ourselves.
A few things became very clear.
First, almost everyone says this looks amazing and they want it. We believe the reason for this is that most people use only a tiny fraction of what AI can actually do.
For regular users, AI still basically means a chatbot.
They are one to two years behind what is already possible. The technical crowd may be building their own workflows and skills, but the majority of people are not there yet.
Second, most people only have a few core workflows they really care about, but they spend hours trying to set them up and optimise them.
They're also spending a huge amount of time just trying to keep up with how fast everything is moving.
And they obviously fail.
Third, almost everyone who works from a computer is trying to force all of this through the chatbot interface, which is incredibly hacky.
What that told us is simple: a vast majority of people do not need deep, complex customization.
They need a solid, simple, out-of-the-box product with a handful of genuinely useful tools and a setup that already works.
So we thought, why not just package that for them?
Our mission is to put what we first built for ourselves, and then for the people around us, into the hands of everyday people. We believe everyone deserves access to this technology, especially in the form that turns out to matter most: a personal assistant.
That is really where our journey began six months ago.
Today, we are building Maevum, a platform that enables anyone to get their own personal AI assistant.
The core problem we are tackling is that today's AI systems do not have a usable working memory.
Technically, some form of memory exists, but in our view, nobody is approaching it the way it needs to be approached.
After speaking with hundreds of people who fit our ICP, we saw the same pattern again and again: every conversation starts from scratch. Every time someone uses ChatGPT or Claude, they have to re-explain who they are, how they work, where their documents live, how they want things done, and how they want tasks completed.
That is where our core differentiation starts.
First, Maevum creates an automatic context layer.
Instead of prompting AI from scratch every session, Maevum continuously builds and maintains your working context across your emails, documents, conversations, files, decisions, relationships, and projects.
It organizes all of that automatically into a synthetic context layer.
Second, Maevum understands not just what you are asking, but why you are asking it.
Most AIs understand: I need to write an email.
Maevum reads that as: you're raising investment, Ricardo is your co-founder, this document is for investors and future employees, and this customer has already seen version two. I need to write an email.
Which one do you think is going to write the better email? A personalized context layer changes the quality of every response.
Third, context management alone is not enough.
A lot of systems stop at organization, but that still leaves the user doing the work.
So Maevum comes with out-of-the-box agentic workflows that actually use that context. Instead of installing OpenClaw, building MCP servers, or configuring workflows from scratch, you sign up, connect your tools, and you're done.
And fourth, Maevum makes context shareable.
Every person's AI gets smarter individually, and every team's AI gets smarter collectively. Instead of just sharing files, you share understanding.
That is the difference for us. We are building an AI operating system that remembers, understands, acts, and collaborates.