Scale
Accel led the Series A in Scale. The WSJ broke the news, and I’m excited to give my perspective on the company.
Scale is at the intersection of two themes we love at Accel, APX and Productized AI. The world wants to be more efficient and software provides enormous leverage for productivity, approaching the limits of physics in storing, transforming, and transmitting information. As a result, the world will asymptote towards software and products will need to be geared towards software developers. Our APX thesis is that more and more business processes will be abstracted into APIs so that developers can easily integrate them directly into their applications, accelerating productivity. AI, like software, will have an enormous effect on productivity. It is still early days for AI, but we believe companies will be able to deliver better products powered by AI. Customers shouldn’t necessarily need to know there is AI at work, just that their experience has been magically improved. Scale hits both points with our simple API and the use of AI in the background to continually improve the value to customers.
When I was a productive member of society [read: software engineer], I used eLance, oDesk, and Mechanical Turk for a variety of tasks including community moderation and training data for machine learning models. I quickly found that those products left a lot to be desired. I had to adjust my own processes for price discovery, quality control, hiring/firing, and application integration as I scaled up my usage. All of that was orthogonal to how I wanted to spend my time!
What I really wanted as a developer was a simple API that I could integrate into whatever application I was working on. I wanted to be able to send a clearly defined task and get a easy-to-use response with guaranteed quality, straight-forward pricing, and a defined turnaround time.
Enter Scale! Scale builds around the customer and would have been a perfect solution to my past software engineering woes. At Scale, we aim to deliver the best combination of user experience, price, quality, turnaround time and scalability for each of our endpoints, which include image annotation, categorization, and audio transcription. Scale consistently and quickly improves those endpoints to get even better for our customers over time. We take a lot of inspiration from Amazon on that front.
I first heard of Scale soon after Justin Kan (I owe you a drink) accidentally launched the company and it got picked up on Product Hunt. After giving Scale a closer look, I reached out to Alex and Lucy, Scale’s founders; Scale was a mid-YC pivot, so there was no data for me to go on.
I quickly realized they were an exceptional team. Alex, an MIT dropout, aced national programming competitions and led an engineering team at Quora. Lucy, a CMU dropout, was a key designer at both Quora and Snapchat. I loved the idea of an API for human intelligence. I decided to lead Scale’s Series A about a month later. Scale really was just two founders, an idea, and a tiny bit of traction at the time. A few months after the round, Alex and Lucy began to work out of my basement in San Francisco.
It has only been ten months since Scale got started, but we’re already doing millions of tasks per month and growing quickly. We work with companies that build self-driving cars, drones, and social networks. We also help companies do a variety of diverse tasks from satellite imaging to sales call analysis. If you think we could be useful for you, sign up and take the APIs for a spin! Scale even gives you some free tasks to get started.
In only ten months, Scale’s team has more than tripled in size (moved out of the basement), and we’re looking to grow a lot more. Check out our Jobs page if this post was interesting to you! I’d be remiss in not highlighting that Scale is an extremely exciting opportunity for anyone looking for a big data set and/or workforce for machine learning or HCI.
