Today we’re introducing our serverless Python backend (and our very belated announcement for our Python blocks). It’s fast, highly available, yet extremely lightweight. We can safely say: it’s going to be a game-changer for you. 🙂

Story time: why we never liked Jupyter

When I was a data scientist at Airbnb, I’d occasionally use Jupyter for long-form analyses. While I tend to be a SQL-first kind of guy, there are certainly some things that are more easily done in Python. So I’d interleave code with narrative markdown cells, and ship analyses straight in Jupyter.

But there were two big problems with this workflow:

1. It felt too heavy. Starting a new notebook always felt like overkill, especially for the smallest bits of work. So I found myself frequently back in the IDE.‍
2. Infra costs were horrendous. We’d requisition on-demand machines, and we’d have to manage them and pay absurd amounts to deal with idling gas usage.

So we decided to tackle these head on.

When we started Hyperquery, we wanted to build the best damn SQL notebook in the world. But over the last couple years, “What about Python?” came often enough that we knew we’d have to dive into the Jupyter world one day. Still, we didn’t want introduction of Python to degrade the delightful user experience we’d agonized over from the get-go — the core of what makes Hyperquery Hyperquery. We were worried that the issues that plague our competitors would come to haunt us. We didn’t want to have to give up our elegant, WYSIWYG editing experience, adding unnecessary friction to the creation process. We didn’t want to have to subsidize the cost of Python by forcing SQL-only shops to foot a heavier bill. We needed infrastructure to remain cost-effective for us and for our users, at scale. We wanted a sustainable, usable solution, not just a throw-money-at-the-problem solution padded by venture capital.

Introducing Python blocks: delightful and serverless

Today, we’re launching our solution, and we think you’re going to like it. We’ve innovated in two big ways here:

1. It’s still WYSIWYG, and it’s lightweight. If you haven’t given it a try already, /python is your gateway into the Python world. And that’s it. Everything will work like magic. We’ve preserved our core elegance and simplicity so you can feel comfortable using Hyperquery for everything, from quick data pull to elegant research findings. But it’s powerful and stateful so you can do your most complex work here as well.

1. We built the world’s first completely serverless Python backend. This means you don’t have to manage infrastructure (if you don’t want to), you don’t have to worry about shutting down idle machines, you don’t have to worry about incurring insane costs. We’ve separated the kernel from the compute, meaning you’ll always pay less on Hyperquery than you will managing your own infrastructure.

Welcome to the future.