Every major technology shift creates a new infrastructure layer. We believe applied intelligence and programmable infrastructure represent the next foundational shift, and we back the founders building it.
The platforms, protocols, and developer tools that form the foundation for the next generation of software. We seek companies reducing the complexity of building, deploying, and operating modern systems. This is the connective tissue of the digital economy, and we believe the winners here will be generational companies.
The modern developer stack is fragmenting at an accelerating rate. What was once a monolithic deployment pipeline has become a complex web of microservices, serverless functions, edge nodes, and multi-cloud orchestration layers. The companies that simplify this complexity without sacrificing power will capture enormous value. We have seen this pattern before: AWS abstracted the data center, Stripe abstracted payments, Twilio abstracted communications. The next wave of abstraction is happening now at the orchestration layer.
We focus on companies with strong developer adoption curves and bottoms-up go-to-market motions. The best infrastructure companies are discovered by engineers solving real problems, not sold by enterprise account executives. We look for products that developers love to use, that spread organically through engineering teams, and that eventually become so embedded in the workflow that they become a line item on the CTO's budget.
Our portfolio companies in this pillar include cloud-native data orchestration platforms, edge computing infrastructure, CI/CD optimization tools, API gateways, serverless workflow engines, and database-as-infrastructure providers. Many of these companies were the first institutional check. We are proud to have backed them when the thesis was still a thesis and not yet a trend.
"The best infrastructure companies don't just build tools. They redefine the boundary between what's possible and what's practical. We look for the founders who see that boundary more clearly than anyone else."
AI and machine learning applied to real-world workflows with measurable ROI. Not intelligence for its own sake, but intelligence embedded in systems where it creates compounding value over time. We are not interested in models as products. We are interested in models as capabilities within products that solve meaningful problems.
The applied intelligence landscape has matured significantly since 2020. The foundation model layer is consolidating around a small number of well-funded players. The real opportunity is in the application layer: companies that take general-purpose intelligence and make it specific, reliable, and measurably valuable for particular workflows. Enterprise decision intelligence, computer vision for manufacturing, ML model optimization, predictive analytics, and conversational AI for specific verticals are all areas where we see extraordinary potential.
We have a particular affinity for teams that combine deep domain expertise with strong ML engineering capability. The best applied AI companies are built by people who understand both the technology and the problem they are solving. A team of brilliant ML researchers building a generic platform will lose to a team of solid ML engineers who deeply understand pharmaceutical supply chains, or manufacturing defect detection, or financial risk modeling.
Our portfolio in this pillar is unified by a common thread: each company has built proprietary data flywheels that make their products better with every customer they serve. These data moats are far more durable than any model architecture advantage. We believe the winners in applied intelligence will be defined by the quality and uniqueness of their data, not the cleverness of their algorithms.
"The companies that will win the applied intelligence era are not the ones building the best models. They are the ones building the best feedback loops between their models and the real world."
As systems grow more complex and interconnected, the attack surface expands. We invest in companies building the security, identity, and compliance infrastructure that makes modern architecture trustworthy. This pillar is not a bet on fear. It is a bet on the structural requirement for trust in every digital interaction.
The security landscape is undergoing a fundamental architectural shift. The perimeter-based model is dead. Zero-trust architectures, decentralized identity systems, and automated compliance engines are replacing the moats-and-castles approach that defined enterprise security for two decades. This shift creates enormous opportunities for companies that rethink security from first principles rather than patching legacy architectures.
We are particularly interested in companies that make security invisible. The best security products are the ones that users and developers never think about. They work in the background, enforcing policies and protecting systems without adding friction to the workflow. This is a design problem as much as it is a technical problem, and we look for teams that understand both dimensions.
Regulatory tailwinds are also significant. GDPR, CCPA, SOC 2, and an expanding patchwork of global data protection regulations are creating structural demand for compliance automation. Companies that can turn regulatory complexity into programmatic simplicity will capture value for decades. Our portfolio in this space ranges from zero-trust access platforms to decentralized identity verification, secrets management, and automated compliance engines.
"Security is not a feature you bolt on. It is an architectural decision that determines whether everything else you build is trustworthy. We invest in the companies that make that decision easier for everyone."
We have backed 68 companies across four funds. Over time, clear patterns have emerged in the founding teams and companies that succeed. These are not rigid rules, but they are strong signals.
We have a strong preference for founding teams where at least one co-founder has deep technical expertise in the domain they are building for. We are looking for people who have felt the pain of the problem firsthand, not people who have observed it from a distance.
The best companies are built on insights that are structurally true, not just timely. We look for founders who can articulate why their opportunity exists now and why it will persist. Timing matters, but durability matters more.
We invest in markets that are large enough to support venture-scale outcomes. We prefer markets that can be clearly defined and measured, even if they are still emerging. We are skeptical of TAM slides that require three assumptions to reach a large number.
Particularly for our infrastructure and developer tools pillar, we favor products with natural bottoms-up adoption curves. Products that individual engineers choose to use, that spread through teams organically, and that eventually become organizational standards.
We look for companies whose products get better with use. Proprietary data flywheels, network effects, and compounding advantages are the most durable competitive moats. We prefer these to IP-based moats, which are often more fragile than founders believe.
We respect capital efficiency without fetishizing frugality. The best founders know exactly when to be lean and when to invest aggressively. We look for teams that can articulate their burn rate in terms of experiments run per dollar, not just runway remaining.
We review every inbound submission. We do not require warm introductions, though they can accelerate the process. Here is exactly what to expect when you pitch Atlas Ventures.
"We started Atlas because we believed the most important technology companies of the next decade would be infrastructure companies. Eight years and four funds later, that conviction has only deepened. The builders of infrastructure are the architects of what comes next."