Backblaze, Inc. Class A Common Stock Investor Day 2026

Good morning, everyone, and welcome to Backblaze's Investor Day for 2026. Today's session is being webcast live, and a replay will be available on our investor relations website following the event. My name's Mimi Kong. I'm the Head of Investor Relations, and on behalf of the team, I want to welcome everyone for joining us today here in person and those tuning in via live webcast. Today we will have a short break around 10:45 A.M. today, just to let everyone know, and we are scheduled to wrap up by noon. Some house cleaning things, housekeeping. Today's presentation will include forward-looking statements, and actual results could differ materially from those statements. Please see the disclaimer on the screen for more details. Full materials, including today's slides, are available on our investor relations website. With that, I'd like to invite Gleb Budman, our CEO and Chairperson, to the stage.

Good morning. My name is Gleb Budman, co-founder, CEO, and Chairperson of Backblaze. Today is our first Investor Day in the five years since we went public. A few people actually asked me, "Why now?" Is there something specific about today that we wanted to share the Investor Day? The reason that we're doing the Investor Day today is because the company and the market have shifted dramatically over the last year, roughly. What I would say is over the 20 years that we've done this, right now is the most exciting time in our journey. We wanted to share that with you. Over the next 2.5 hours, this is the plan. I'm going to talk about how the market has shifted and what our role in that is.

Dan Spraggins, our CTO, is going to talk about the platform and why it's uniquely situated for AI. Anuj Kumar, our CRO, is going to talk about the signals that we're seeing with AI and how we're building that into a repeatable and scalable engine. Marc Suidan, our CFO, is going to talk about how we're changing this opportunity into a great business. We're also going to hear from Hume AI, a leading multimodal AI company, and we're going to hear from WEKA, a new partner of ours, in the high-speed storage space. Listen to how those all come together into stitching together this ecosystem of AI.

We're going to cover a lot of information today, but if there's one thing that I'd want you to walk away with, it's that not only is AI creating a ton of data and using a ton of data, but the role of data and data storage is fundamentally changing, and that's how we're going to talk through it today. AI is creating this seat at the table for an independent capacity tier of storage. We're going to explain why that is and what we mean by an independent capacity tier of storage. We believe that that seat is open, and we believe that seat is for us. Part of the way that we are ending up there is we started this company 20 years ago. We started it solving a problem that was really important at that point, which was backing up computers.

We planned to do that on Amazon S3 and store the data there. We worried the economics were not going to work. So we ended up building a full stack solution: servers, a cloud storage file system, and the operational expertise to manage all that because we had to store massive amounts of data and do that very efficiently. So in doing that, we then spent 10 years optimizing that full storage stack. We then launched it out to developers and enterprises as B2 Cloud Storage. Over the last 10 years, we've been optimizing and scaling that platform, and today it's ideal for the AI use cases that are being needed in the market. So that's the foundation from which we believe we get to have this seat at the table as an independent capacity tier.

The rules of storage are being rewritten with AI, and here are the three rules that are being rewritten. Number one, data was growing steadily for years and years and years. Basically always, right? There's now this step change that is happening where the data is just exploding, both in terms of the amount of data being created and the way it's being used. One of the interesting things with it is that it's not just the largest companies anymore. Even small new companies have massive data sets and massive data needs. That is a significant shift from the past. Number two is there used to be three hyperscalers, and companies defaulted to one of them. Now there are about 200 AI infrastructure companies out there, and that is completely shifting how companies are building technology.

Number three is that storage used to be just a thing companies used, but today it's a critical performance layer of whether companies can innovate in AI and whether they can get an ROI from AI. Those three rules play perfectly into Backblaze's strengths. As these data sets scale massively, they need to go somewhere, and they need to be efficient in where that storage is put. Backblaze's optimization of our platform for the last 20 years makes it possible to store massive amounts of data and to do that efficiently. The explosion from three hyperscalers to 200 neo clouds has two benefits where Backblaze plays in. One is that those AI infrastructure companies themselves need cloud storage as part of their workflows, and we're ideally situated to provide that to them.

The other is that all of the AI natives and AI builders using all of these companies need an independent capacity tier to store the data to be able to use them. The third strength is that as companies care about the performance per dollar of their storage system, because it is what's critical to get both the innovation out of AI and the ROI out of AI, Backblaze's optimization of that exact thing from hard drives over 20 years is exactly what they need to be able to do that. So this all supports us taking advantage of what is a large and growing data set. The data explosion over the next few years is hard to fully wrap our heads around. Right? In 2024, IDC said there was 173 zettabytes, created.

In four years, that number is going to quadruple to 700 zettabytes created in one year. Not all of that data is going to be stored. Not all of that data is going to be used. But an increasing amount of it is, because AI is also making it possible to get value out of the data that is getting created in ways that was not possible before. Not only is the amount of data getting created going exponential, but the desire to keep and use it is as well. That supports a large and fast-growing market that we get to participate in. We continue to serve the markets we have always served, the core markets where we have served disaster recovery, backup, archive, media workflows, application storage. All of those markets still exist and are growing.

That has not gone away, and that is still an important part of the trajectory of our business. AI layers on a big market on top of that and a faster-growing market on top of that. That requires a company that can scale with that. That is what we have built. We now have over 5 exabytes of data storage under management. That makes us one of the largest cloud storage companies on the face of the planet outside of the hyperscalers. We are planning to expand that capacity by about 30% this coming year. Obviously, that takes hard drives, networking, servers, data center space, power to do all of that, and having the ability to acquire that and the relationships to do that takes a certain level of expertise. On top of that, it is the actual managed storage aspect of it that is really hard.

It is the software platform that can scale like that. It is the operational expertise and rigor that has been built around that to be able to actually deploy and manage scale at this rate and this scalability. We have talked about the scale of the market. Let us talk about the second thing that we mentioned, how the rules are changing. The dispersion of these AI infrastructure companies. This is probably the most important slide that I am going to show you, and so I want to spend a little bit of time on it. For the last 20 years, technology was built inside of a hyperscaler. Generally one hyperscaler. Even when there were three hyperscalers, companies would pick one and use it and build all of their technology stack inside of that one hyperscaler.

Maybe they were inside of Microsoft, maybe they were inside of Amazon, maybe they were inside of Google, but they picked one, and they built everything inside of it. Today, there are 200 AI infrastructure companies, neo clouds, sovereign clouds, inference clouds. That is not even starting to talk about the fact that all of these clouds have different regions, including the hyperscalers. A common customer has their base workloads inside of one of the hyperscalers still. They are using databases, maybe they are using networking, maybe they are using 1 of the 200 different services, or 10 of the 200 different services that the hyperscalers provide. They are still using a hyperscaler for a bunch of stuff.

They go and they say, "Now I need to do model training." The hyperscalers don't have access necessarily to the latest chips or have availability of them, or they don't have the price point of them. So they go to one of the Neoclouds and they say, "Okay, I'm going to use you for the model training piece of it." That one Neocloud doesn't have everything they need, so they go to another one for some of their other model training aspects. So now they've got base workloads in a hyperscaler, data they're collecting somewhere for the model training, and then using two Neoclouds for model training. Once they've trained their model, then they want to do inference somewhere where it's optimized for that use case. So they may be using two or three different inferencing providers around the world.

Beyond that, then they may also be using different regions of these. What that means is that as that AI builder, you now have four, five, six, eight places where your data needs to go. If your data is sitting inside of one of the hyperscalers, you're paying massive egress fees every single time you send your data to any of these places. So you really need this independent capacity tier, a place to keep your data so that you can innovate with AI. It's not just about saving money, it's about the fact that you want to use the AI infrastructure that's out there in order to innovate.