
Building with AI Agents
I saw a video of an AI agent that could actually do things. Not just talk. Do. That was March 2025. I have been building with Manus ever since.
In March 2025, I came across a video of Manus AI Agent doing something I had not seen before. It was not just generating text or answering questions. It was actually building things: researching, coding, designing, deploying. The agent was completing tasks that would take a person hours, and it was doing them end to end without hand-holding.
I signed up immediately. Within weeks, I was using Manus to build projects I had been putting off for months. A comparison tool for AI models. A language preservation platform. Educational resources. The speed was unlike anything I had experienced with other AI tools. This was not a chatbot. This was a collaborator.
By May 2025, I was not just using Manus. I was hosting events, introducing it to communities, and showing people what becomes possible when AI can actually execute, not just suggest. I became one of the first Manus Fellows, and I started organizing workshops and celebrations across the Northeast.
For fifteen years, I have been working on projects where the bottleneck was always the same: not enough hands. Wikipedia articles that needed translating into dozens of languages. Healthcare information that needed to reach rural communities. Educational tools that needed building but had no engineering team behind them.
Manus changed the equation. Suddenly, one person with a clear idea could build what used to require a team. I built this very portfolio website with Manus. I built AI Side by Side, a tool for comparing AI models, in a single afternoon. The technology aligned perfectly with everything I had been trying to do: make knowledge accessible, bridge language barriers, and put tools in the hands of people who have ideas but not resources.
"AI should not just think. It should also do. That is the difference between a chatbot and a collaborator.
From the very first Manus event at Harvard to community celebrations across the Northeast, I have been bringing people together to experience what AI agents can actually do. These are not product demos. They are spaces where people build real things, share real stories, and leave with something they made.
The very first Manus event, held at Harvard. We brought together high schoolers and seasoned professionals, non-technical newcomers and AI product engineers, all in one room. Participants went through live demos of Manus, then formed teams to co-design AI-powered educational tools. By the end of the afternoon, groups had prototyped an evidence-based learning platform for diverse learning needs, a social-emotional learning companion, a job displacement analysis tool, AI-generated learning videos, and augmented reality experiences for physics.

A workshop at the public library that blended live demos with a build-and-share sprint. The room was a mix of first-time job seekers and retirees, and that was the point. Several retired attendees said Manus made it feel realistic to launch a venture at their age. Three new business ideas took shape on the spot, including a local tour service and a custom cookie shop. People used Manus to build websites, rewrite resumes, clean up spreadsheets, and run quick market research. Five participants upgraded to Manus Pro before leaving, and library staff asked about hosting more sessions.

When Manus turned one, ten cities around the world celebrated. I organized the Boston event at Shake Shack in Harvard Square. Builders, researchers, founders, and the AI-curious gathered for an evening of live demos, a buildathon, and real stories from people using Manus in their work. Between bites of burgers and rounds of foosball, people were shipping actual projects: educational tools, personal dashboards, and more. Strangers left as friends. That is what community looks like.


A comparison tool that lets users evaluate AI models side by side with real-time responses. Built in a single afternoon with Manus to help people understand the differences between models and make informed choices.
aimodels.manus.spaceThe event website for the Manus 1st Birthday Celebration in Boston. Built entirely with Manus, featuring event details, the Year One story, and a love letter to why Boston was the perfect city for this celebration.
boston.manus.spaceThis portfolio website was also built with Manus. Every page, every animation, every line of code. It is a living example of what becomes possible when AI can execute, not just suggest.
Manus is the tool that ties my other work together. The same technology that powers this portfolio also helps me build language preservation platforms, educational tools, and healthcare resources at a speed that was not possible before.
Manus is part of a broader journey in AI, alongside Pickaxe AI and academic research at UVA.
The same drive to make knowledge accessible that led me to Wikipedia now fuels my work with AI agents.
Manus events share the TED spirit: bringing people together around ideas worth spreading.
AI tools like Manus accelerate healthcare content translation and accessibility work.