A Contain(er)ed World
I spend most of my time looking for code that already accomplishes what we need to do, or rather finding the closest thing possible, adding our 20%, and moving on.
Young companies look to open source for inspiration, tooling, and small pieces of code that do the one thing they need it to. You combine open source code with your own proprietary code and through combinatorial innovation, a “disruptive” startup comes out the other side.
This is even true for “ex-Google, MIT PhD starts new company”. Let’s say for example this new company is called X. X does machine learning for Y. For example, X uses mammograms to detect breast cancer with higher accuracy. This is amazing!
How much of that code is from the all-star PhD? It’s most likely the core algorithm. Okay great, but that’s not going to be enough to sell to an enterprise. It’s difficult to sell an SDK to a hospital. They want a UI, they want a backend, authentication, RBAC, reporting, perhaps a mobile app, etc.
Let’s say generously the machine learning algorithm is 20% of the code. That other 80% is standard web dev stuff:
- The backend is some open source DB
- The front end is React / Vue, etc.
- The backend is Node, Golang, etc.
Digging in further, what are the basic components of an enterprise-ready app?
- Product Assortment
- Single Sign On
- Audit Logs
- Role Based Access Control
- Change Management
- Product Security
And that’s just on the product side, let alone all the other pieces that make a company a company. My point is that most of the solutions to the basic components of an enterprise app are the same. No matter how transformative the product, people still need to log in, sign up, learn how to use the product, etc.
My hope is that the future of enterprise product development becomes faster, which is why I built SugarKubes. Imagine being able to pull down code or containers that satisfy each of these requirements without having to have a large dev team? I didn’t find it out in the market to my liking, and so I built it.
