AI Agents: A 30-Day Public Journey to Building an Intelligent Data Visualisation Agent 🚀

Have you ever heard of an AI Agent 🤔? Great, not many people have either (at least not yet). But trust me, you’re about to.

But the first question is, what actually are AI Agents? In this article, I’m going to explain that and how we’re going to build one ourselves, from scratch.

This article is the first out of a 30 article series, were we document the process of building an AI agent product within 30 days. Check it all out below 👇.

The Next Paradigm Shift

Well, for the past 18 months, we’ve embraced a world where AI co-pilots have assisted humans with tasks ranging from customer service to software development 🤝. AI has truly become our productivity-enhancing companion. But the technology has advanced at such a rapid pace that co-pilots are now somewhat outdated 🤯. The paradigm has shifted from humans doing most of the work with AI assistance, to AI being able to complete end-to-end tasks, with humans simply assisting with the output ☄️. The game is changing, and up steps the agentic workflow, the new wave of intelligent computing.

What are Agents

AI Agents are autonomous systems or programs that can perform tasks, reason, make decisions, and interact with users or other systems in a similar way to humans 😅. Is that it? Well, no. If you ask an agent to do something, it has the ability to plan the tasks that it needs to do in order to achieve the goal you set out 🥅. Then it will leverage the tools it has access to in order to achieve each task step by step.

For example; if you ask What is the weather in London tomorrow🌂? An agent with access to a search tool will search the weather reports on the internet, extract them and summarise the weather report in London for you 👀.

The Challenge

Over the next 30 days, we’re gonna be designing an AI agent, packaging it into a super user-friendly product, and then — get this — releasing it as open source 🥫. You’re probably wondering “why open source?” Well, we’re super passionate about making AI accessible to everyone, and we truly believe that by doing it this way, we can create some serious value and impact the lives of a huge community of tech enthusiasts and pros 🤝.

The Goal

So, we wanted our agent to automate and make an impact in an area that is tough to do, and turn it into a piece of cake 🍰. A lot of our previous work in Cubode was around making data visualisation more intuitive and easier for non-technical users to understand, so we wanted to build on that momentum. Now, data visualisation is nothing new, but the whole idea of using AI to automate the creation of interactive charts that can be embedded elsewhere, without any coding? That is problem we know we can tackle 🙌. So here’s what we’re hoping our agent and the final product will do for everyone.

1. Receive users uploaded their data 💽
2. Auto generate interactive data visualisation charts and provide them to the user 📊
3. Enable the user to have full control over chart customization 🚘
4. Enable the user to export and integrate the chart into their current workflows, whether as images for presentations or embedded in Notion ✈️

How are we going to do it?

We’ve split the project into 3 tech sections that really show off what our team can do, and will be the backbone to the final product 🧠. 30 days to pull this off? Yeah, it’s a big ask 🙃. But hey, we’re not the type to back down from a challenge. Here’s how we have structured the project, and the articles over the next 30 days will be focused around our work in progress within either of these sections, so get ready 🎯!

1. Front End Development 👨‍💻
2. AI Engineering ⚙️
3. Back End Development 👩‍💻

Join Us

Stay tuned, we will release an article every day for the next 30 days with updates on our progress 🚀.

If you are interested in being one of our first users of this agent then go to www.cubode.com to sign up to get the first access.