Tecton.ai announces $35M Series B led by Andreessen and Sequoia

When it comes to amazing VC firms you want backing your startup, two of the most common companies that come to mind are Andreessen Horowitz and Sequoia Capital. This week Tecton.ai announced a $35M Series B round led by both firms and if past trends prove true, it’s likely hundreds of millions of dollars of funding are ahead for Tecton as it continues to pioneer in the machine learning world.

I am personally super excited for Tecton.ai because they provide a service that is very relevant to what I do every day which is deploying machine learning models at scale. While machine learning gets tossed around a lot as a buzz word, how it works under the hood is often not very well understood.

While I’m not going to try to explain the entire machine learning ecosystem to you in one blog post, I did thought it would be interesting to discuss a bit more about what Tecton.ai does and why it’s a big deal.

At the core of machine learning is the concept of features and labels. I always like to think of features as column headings in spreadsheets. Let’s imagine that I want to build a machine learning model to predict the best time to go hiking in Yosemite. I’d probably want to understand things like traffic, weather, # of people in the park, and the number of hiking trails that are open. These things I just listed off, those are features, they are the fields used as input by your machine learning models. The labels are the output, i.e. the data that goes under each of the column headings.

Before machine learning existed I would have to pour through spreadsheets of data and try to manually, or though some wacky formula, determine the most optimized solution. Then, as things change and new data is added or removed I would need to update my formulas, I had to do the learning 🧠

Today, machines can do the learning but one of the key ways to optimize how they learn is to have the best features and the cleanest data feeding into the system to make sure you’re making models based on signal, and not noise. Tecton.ai makes it easy to build feature from a myriad of sources and deliver them into production instantly, which trust me, is a pain in the you know what to do manually as most ML companies do. Here’s an overview of the solution:


As you can tell, I’m pretty excited about what Tecton.ai is doing but of course I couldn’t help myself from looking at what was being done with the .COM…

Tecton Corporation

Yup, it’s an old broken site that if you look at the very bottom right corner you’ll find a little button that says “enter.” Clicking that takes you to this website…which made me feel like I was back in the early days of the interwebs.

Tecton Corporation

I’m not sure if Tecton.ai would ever really want to make a run at the .COM even though it looks like it was probably abandoned some time ago. The reality is, Tecton.ai’s target customer is data scientists and the .AI extension definitely speaks to this crowd more than a .COM. Given that Tecton.ai has already raised $60M it’s pretty clear that the .AI extension certainly doesn’t seem to be holding them back with customers or investors.

Congrats to Mike, Kevin and the whole Tecton.ai team, not a bad way to end the year! 🎉

Morgan Linton

Morgan Linton