It’s hard to think of a better time to be a developer than today: you can get a frontend, a database, and an app server up and running in literally minutes. But at the same time, the apps we’re building are getting a lot more complicated: they’re bigger, they’re integrating real time audio and video, they’re streaming, and they’re relying on data – ML or otherwise – to make decisions. Progress in developer productivity and infra over the past decade has been exciting, but we’re just getting started.
After what feels like a decade of slow but sure progress, ML and AI are finally making it into the mainstream through LLMs, foundation models, and the like. For long time practitioners who have been grinding for years, it’s an exciting time. It seems clear that everything we do and use will have some sort of ML behind it – but less clear how that happens in practice. And to get there, we’ll need a supporting cast of developer tools, data and backend infrastructure, and analytics to emerge.
Before Hex, dbt, and Snowflake we had Tableau, Kimball, and Hadoop. The principles of the Modern Data Stack are similar, though: help teams make quick, effective, trustworthy decisions with their data. And yet, 10 or so years into this trend, most data teams will still tell you that they’re not having the impact they’d hoped for. We’re excited about tools and infrastructure that enable data teams to guide decision making, share knowledge, and support better user experiences.
As the wave of digitization washes over every industry and the line between the physical and digital world blurs, new attack vectors are leaving businesses more vulnerable than ever. Protecting infrastructure, end-user devices, the software supply chain, and everything in between has never been more pressing for CISOs. With software touching and transforming every part of the enterprise, security can no longer be something organizations bolt-on but needs to be tightly coupled with the entire engineering and IT org. We believe there is a massive opportunity to improve security by integrating best practices and prophylactics directly into developer tooling and the software delivery cycle. New primitives will be necessary to ensure the security of robotics and real-time decisioning systems that live amongst us. Finally, we expect leaps forward in machine learning, cryptography, and advanced threat hunting techniques to drive an innovation cycle for a new class of products that help thwart an ever-growing set of sophisticated attacks and attackers.