How to build an interactive biosecurity tool with minimal coding experience
Want to build an interactive tool for your biosecurity project but feel stuck because you can’t code? Here’s how you can make it possible in just one day.
I recently built an MVP clinical metagenomic sequencing simulator in just one day using AI tools — something that would have previously taken me several days. So, if you’re dreading the thought of coding, this guide will show you how to leverage AI to build an interactive tool faster than you thought possible. Even if you didn’t think you could code it!
After reading this, you'll be able to:
- Plan and build a basic interactive tool
- Set up your AI-assisted coding environment
- Avoid common pitfalls in using AI tools for coding
But I don’t have the coding skills…
Let’s test your coding knowledge. Take a look at this simple code snippet:
If you can understand that this code counts from 1 to 5, you're more than ready to build your first tool. If not, I’ll recommend a different coding tool. Stick with me!
The AI will handle the complex parts — you just need to understand the basics.
Before you start building, plan out your tool
While it is tempting to build something quickly, Resist the urge!
Spend at least 1 hour planning out what you're trying to build. The clearer and more detailed your plan, the better the AI will execute this plan for you.
The AI is an assistant. It can't read your mind. So if you are confused about what you want to build, the AI's output will be equally confused.
Consider the following:
- Why are you building this interactive tool?
- Who is this interactive tool for?
- How does your tool solve a challenge for your target audience?
- Mockup your interface on paper, Canva or a UX tool like Figma
- Map out the steps your user will take when using the tool
Set up your AI-assisted coding environment
bolt.new
If you’re fairly new to coding or have a fear of the terminal, I’d recommend starting with bolt.new. Prompt it based on the plan you came up with and upload images of your mockup.
For example, “Build me a tool for [target audience] to [purpose of tool] based on the mockups I uploaded.”
It will build and run the tool for you, so you can start using it instantly!
Cursor
If you have some experience with coding or would like to dive deeper into coding, then I would recommend using Cursor.
- Follow this tutorial
Learn how to download Cursor and get a sense of how it works. - Try these three exercises:
- Prompt Cursor to build something simple. (aim for 1-2 sentences!)
- Highlight a line of code and ask “What does this do?”
- Type @Codebase in chat and ask “How can I improve this?”
- Deploy the app in your local environment (if you don’t know how, ask Cursor!)
I’m sure there are other AI tools out there that might be better than the two I’ve recommended so far, so I’d encourage you to have a go at building with one of these tools or looking up others.
Build a basic interactive tool
Start with the core feature
When building my metagenomic simulator, I started with the prompt:
"Build a Streamlit app showing infected people over time, with 5 contacts per person and 0.1 transmission probability."
Notice that I didn’t include any other more complex features like randomly generating the number of contacts per person or defining a function to vary transmission probability.
Start simple and add complexity as you go along because it’s important to…
Understand the suggested code before applying it
While it is tempting to go on autopilot and accept all the AI’s suggestions, your code will quickly get extremely complex and error-ridden.
Try understanding the code it has suggested or ask the AI to explain it to you before moving forward.
Going slow will give you more confidence in how your tool works. It will also help you learn what the code is actually doing, allowing you to spot errors and give better suggestions for improvement.
Save your changes
Nothing is more painful than making changes that break your code and not knowing how to reverse them.
To avoid this, push your code to GitHub regularly. Ask Cursor, “How can I save my code to GitHub?”
The best practice is to save your work after every major milestone. For example, if you’ve just added a working button to your interface, push the changes before moving on.
Share and get feedback on your work
One of the best things about interactive tools is sharing them with others! Once your tool is ready, use Cursor to deploy it online so anyone can try it.
Ready to start?
Building an interactive tool doesn’t have to be intimidating. With AI assistants like Cursor, Bolt, or Microsoft Copilot, coding is now more accessible than ever. Start simple, plan thoroughly, and take it step by step.
We’re excited to see what you end up building!