From first line of code to App store submission in seven days: How Alumworks used Gadget to ensure their AI-powered product was first to market
The Problem
After years of working at Shopify listening to all the struggles merchants deal with day in and day out, three ex-Shopify employees — Jason, Josh, and Ryan — set out to solve those problems, founding their agency Alumworks in the process.
The plan was simple: build a handful of highly-differentiated public apps to simplify store operations for Shopify merchants.
But the success of one of their early apps, Alpaca, coupled with their experience working at Shopify, brought Alumworks to the attention of merchants and partners looking for custom apps.
[Gadget] just handles so much of the boilerplate and the busywork that we can easily triple or quadruple our output.
- Jason Normore, CTO @ Alumworks
Despite the change in direction, custom apps would unlock a new revenue stream for the business, and so Alumworks jumped in wholeheartedly. They picked up a few custom projects and completed them in no time. But what they didn’t anticipate was suddenly a huge chunk of their time was consumed by maintenance for their ever-growing fleet of custom apps. Limited by their own rapid success, they had less and less time to dedicate to exploring new ideas and opportunities to create their own proprietary apps.
“Shopify apps require a fair bit of maintenance, even after they’re done being built: you must constantly tweak your app’s infrastructure to ensure that it can scale to meet the needs of a growing user base, you have to deal with bugs and customer support issues, and you need to stay up to date with the latest Shopify API changes and annual upgrade cycles. We’re a small team and this all takes time. Time that we could be investing in building new applications.“ - Jason, CTO @ Alumworks
In late 2022, Alumworks finally had an app idea that couldn’t wait. The launch of ChatGPT demonstrated how powerful AI-generated content could be, and the team knew they wanted to be the ones to bring it to merchants, so they had to move fast.
They came up with the concept for Otto, a GPT-3-based content generator to help merchants create SEO-friendly product descriptions and blog posts. With the immense amount of interest in generative AI solutions, getting their product to market first was imperative.
The Alumworks team needed a platform that could accelerate their work, and give them an edge over their competitors. And they found it in Gadget.
The Solution
The Alumworks team first heard about Gadget through the Shopify network, as both companies were founded by ex-Shopify employees. They were immediately drawn to the value of Gadget. The product promised to cut down on the time and effort their developers spent on each project by handling the boilerplate code and infrastructure that often make up more than half of the work involved in building Shopify apps.
A single tool that could help us build out new projects faster, while also handling hosting, scaling, and Shopify’s annual API version updates was exactly what we needed to continue growing our business, while staying small and nimble.
- Jason Normore, CTO @ Alumworks
On top of that, Gadget’s serverless infrastructure and built-in Shopify API connection meant that there was very little ongoing technical work required to maintain apps, something they were still frustrated with from their time building custom apps. They were sold — almost.
“Gadget’s value proposition really resonated with us. A single tool that could help us build out new projects faster, while also handling hosting, scaling, and Shopify’s annual API version updates was exactly what we needed to continue growing our business, while staying small and nimble.”
While excited about the prospects of the product, the Alumworks team was understandably hesitant to bet the stellar reputation they had built, on a technology that they had just recently discovered. They had to test Gadget to make sure it would deliver on all of the lofty promises it made. So Alumworks sent in Jason, their CTO, to try Gadget on a few small projects.
Jason explained, “As a backend developer with decades of experience and my own aesthetic for how I like to build software, I was initially skeptical of Gadget’s claim that it could greatly accelerate development time on any project, regardless of the use case.”
That skepticism didn’t last long as Jason began to explore Gadget.
“I tried Gadget out on a wide set of small sample projects designed to test out the limitations of their framework. I was pleasantly surprised that Gadget could handle all of my requirements, while also reducing the amount of code I had to write by at least 50-70%,” Jason shared.
Cutting out 50-70% of the code also meant cutting down on all the time he would have spent researching, troubleshooting, and rewriting that code. They didn’t have to worry about OAUTH, API connections, or databases, they only focused on building their app’s unique logic. The time savings were immeasurable.
Once Jason was confident that Gadget’s infrastructure, and framework, were up to Alumworks’ standards, the team dove in and began building Otto.
The product required the ability to fetch large amounts of data from Shopify and pass it over to the GPT-3 APIs to generate SEO-optimized content, which would then be displayed to the merchant for approval. Once approved, the content would be programmatically uploaded to the merchant’s Shopify store.
With Gadget’s built-in Shopify connection, there was no need to build OAuth, webhook processing pipelines, or the ability to synchronize large data sets. Alumworks simply picked the Shopify API scopes and models needed to power their application, and Gadget handled the rest.
“Gadget’s built-in Shopify connection was a huge timesaver. We had Oauth up and running, and were syncing data to our Gadget backend within minutes. This would normally take us weeks, or even months to build out on our own.”
From there, it was a simple matter of transporting the information to GPT-3 in order to generate the desired content. Because Gadget supports installing Node.js packages, integrating with OpenAI's NPM package took no time at all. With Gadget handling all of the boilerplate, all Alumworks had to do was write a few JS functions.
“The Gadget framework is incredibly well thought out. It handles so much of the boilerplate work for you. I just had to write a few functions, and that was it for the entire backend. I instantly had a fully productionized API, and my own React client, and could immediately begin working on our frontend. We finished the entire project in about a week — from first line of code, to app store submission.”
The Result
With their app submitted in no time, the Alumworks team was able to get Otto onto the Shopify app store well before a plethora of competing apps appeared. During the ensuing months, they went from 0 to over 500 installs, with very little competition. They even found themselves featured on the Shopify App store.
The added visibility increased app installs ten-fold overnight, which was handled without a hiccup on Gadget’s serverless infrastructure.
“The app kept growing, but we didn’t spend a minute thinking about scaling our infrastructure. Everything just kept purring along, regardless of the throughput.”
With a successful launch behind them, the Alumworks team quickly realized that Gadget could be a permanent fixture in their technology stack, allowing them to increase their team’s output, and reduce the maintenance burden of running a fleet of apps.
“We’re fully bought into Gadget now. If we’re going to keep our team small, while juggling custom app development alongside building our own products, we need Gadget. It just handles so much of the boilerplate and the busywork that we can easily triple or quadruple our output. We’re even considering porting our old apps onto Gadget to get all of the advantages of their infrastructure and framework.”
Less code, less maintenance, and more time means that with Gadget, Alumworks isn’t worried about the increased workload from their custom app work. They’re confident they can balance supporting their existing customers, while also creating their own proprietary apps the way they always planned.