Legrande Health Web App
2025-09-09
I've only been working on this project for about a month, but I'll come back and update this post as I undoubtedly learn more.
I recently started working on a web application for Legrande Health. It's an order management platform for medical practices and suppliers. The application is built using React for the frontend, and FastAPI in Python for the backend. The database is managed with PostgreSQL. Frontend is hosted statically in Firebase hosting, and the backend is hosted on Google Cloud Run using Docker containers.
A few things I have learned in the just over a month that I have been working on this project.
The XY Problem
This is a product that is used both internally by the Legrande team, and externally by their customers. This means that there are a lot of different stakeholders with different needs and requirements. I have learned to ask a lot of questions and to really understand the problem before jumping into a solution. This has helped me to avoid the XY problem, where I try to solve a problem that is not the real problem.
It is definitely true that users will come with a solution in mind, but it is my job to understand the problem and to find the best solution for it. This has been a great learning experience for me.
Shipping Early and Often
This is a product that is being built from the ground up. There is no existing codebase to work with, and there are a lot of unknowns. I have learned to ship early and often, even if the product is not perfect. This has helped me to get feedback from users and to iterate quickly.
To be clear, we don't ship code that is not ready for production. But we have a bias towards shipping features that are good enough for the user to start using them, and then iterating on them based on feedback. This has been extremely beneficial for the project.
AI is not Competition
When I started my degree in computer science, AI was just becoming more prolific, and I felt like i had to compete with it. I felt like I had to be better than AI, and that I had to learn everything about programming to be able to compete with it.
For the first year and a bit of my degree, that was a beneficial mindset because it pushed me to learn a lot. But as I have progressed in my career, I have realized that AI is not competition. AI is a tool that can help me to be more productive and to write better code. I have learned to embrace AI and to use it to my advantage.
AI is not going to replace programmers, but programmers who use AI will replace those who don't.
I've seen this quoted multiple times on LinkedIn, and I think it's very true. I don't think that in the near future AI will be able to replace programmers, but I do think that AI will be able to help programmers to be more productive and to write better code. Programmers who use it will surely replace those who don't.
With that the AI's that I use most are Claude Sonnet 4, and Google Gemini 2.5 Pro. I like their CLI interfaces, but I mostly use them through the Copilot extension in VSCode. Using these has allowed me to come into a new codebase and be productive very quickly. I can ask it questions about the codebase, and it can help me to understand it better. It has also helped me to write and debug a lot faster than I otherwise would have been able to do in my first month.
Conclusion
This has been a great learning experience for me. I have learned a lot about building a product from the ground up, and I have learned a lot about working with different stakeholders. I am excited to continue working on this project and to see where it goes.
Fun Fact: The internal name we use for this project is "nomp" which stands for "new order management portal".