BuilderLab

Inspiration

As engineering students whenever we had to create hardware for our projects we were frustrated by the time consuming, expensive and complex process behind it. Hence we decided to find a solution which consisted of creating a software solution that made making hardware designs easier and more efficient. The goal was to enable founders, engineers, and students to create manufacturable designs without having a prerequisite of in-depth technical knowledge.

What it does

BuilderLab is a one-stop platform that bridges CAD software and manufacturing for builders who now can transform their ideas into reality with just a simple prompt. Using this prompt the platform would be able to generate 3D models, hardware, and other relevant programming specifics for that particular idea with the help of various AI APIs. After this process, users can send these designs out to our partnered manufacturers such as Xometry to manufacture the physical product.

How we built it

We built Builderlab using various AI models such as Claude for reasoning and programming, KCL and our self-trained AI for 3d modelling. The online platform was created using HTML for its frontend, and JavaScript for the backend.

Challenges we ran into

Our self-trained AI can sometimes confuse itself and not generate the relevant 3D models. To fix this issue we had to continuously monitor and train our AI so that they can become more accurate.

Accomplishments that we're proud of

China Global Competition, finalist for Chua Thian Poh Pinnacle Prize.

What we learned

Teamwork
Programming
AI systems

What's next for BuilderLab

We plan to scale up our startup to evolve from an AI wrapper company to one that uses its own self-trained models that have been customised specifically for 3D generation and programming. We also aim to organise our own hackathons, and partner with various incubators to help fellow startups in their prototyping and manufacturing journey.