Konbini - Development Log #1

I'm a few weeks into this project now, so I guess it's about time to start Building in Public and talk a bit about what I'm doing.

Long story short, it's called Konbini, and it's a platform you can use to connect to many different ecommerce platforms using the same API -- if you've heard of Plaid, then what I'm trying to do is a lot like Plaid, but for ecommerce rather than banking.

Today, as an ecommerce app developer, you have the choice between either going all-in on one platform, or spending absurd amounts of time writing integration code for every platform you want to support. Konbini's value proposition is allowing you to support a multitude of platforms while only needing to write code for one (modern, developer-friendly) API.

This is a bit of a hot space right now for a few potential reasons (the ecommerce boom fueled by the pandemic, more and more big ecommerce platforms adding APIs as a standard feature in recent years, etc.), but the competition seems to be all either clearly pretty early days, or pretty different from what I'm trying to build:

• Weav - YC-funded, acquired earlier this year, potentially no longer relevant?
• Rutter - YC-funded, looks very similar in concept, no public pricing or signup
• API2Cart - small Ukrainian company, branding and API docs don't really inspire confidence, product seems solid but simpler than what I'm going for. Pricing is public, but pretty expensive ($500/mo for the lowest tier)

The Stack

Code-wise, the stack I'm using for this is full-stack TypeScript, partly because it's a stack I already know and like, but also because I think it's pretty much the ideal stack for a one-man project like this. TypeScript is pretty light on the ceremony, few open-source package ecosystems can compete with JavaScript's, I can effortlessly share code between different parts of the codebase (including the frontend), and there's none of the overhead that comes with context-switching between different languages.

Deployment-wise, I'm looking to (a) take advantage of the most cutting-edge tech, (b) not spend any money I don't need to spend, and (c) give myself as little infrastructure to manage as possible. Serverless tech ticks all of those boxes, so the stack I'm looking at is something along the lines of:

• Google Cloud Run - primary serverless runtime
• Google Cloud Pub/Sub - for queues and asynchronous communication between services
• PlanetScale - serverless, effortlessly scalable SQL database platform
• Cloudflare Workers - for any backend code light enough to run at the edge without access to Node.js APIs

The Roadmap

The above is the current state of the launch roadmap for Konbini, which is actually a backwards plan, a technique I learned about from @amyhoy of Stacking the Bricks. Green blocks are "done", yellow ones are "in progress", and red ones are things I haven't started working on yet.

While a marketing site isn't something I'm strictly going to need for a while, I'm hoping to at least have a simple "coming soon" page up soon, so I have something to link people to, and somewhere people can sign up to be notified when we do get closer to the launch date.

That's all I have for today. Until next time!