Need smarter insights in your inbox? Sign up for our weekly newsletter to get only the things that matter to enterprise AI, data and security leaders. Subscribe now
E2B is a startup that provides cloud infrastructure specifically designed for artificial intelligence agents, and has taken advantage of the surge in corporate demand for AI automation tools to shut down a $21 million Series A funding round led by Insight Partners.
The funds highlight the rapid adoption of AI agent technology, as 88% of Fortune 100 companies have already signed up to use the E2B platform. The round included participation from existing investors Decibel, Sunflower Capital and Kaya, along with prominent angels including Scott Johnston, former CEO of Docker.
E2B technology addresses critical infrastructure gaps as businesses deploy more and more AI agents. It is an autonomous software program that can perform complex multi-step tasks such as code generation, data analysis, and web browsing. Unlike traditional cloud computing, designed for human users, E2B offers a secure, isolated computing environment where AI agents can safely execute potentially dangerous code without compromising enterprise systems.
Vasek Mlejnsky, co-founder and CEO of E2B, said in an exclusive interview with VentureBeat: “E2B solves this by equipping AI agents with a secure, scalable, high-performance cloud infrastructure designed specifically for the deployment of production-scale agents.”
The AI Impact Series returns to San Francisco – August 5th
The next phase of AI is here – Are you ready? Join Block, GSK and SAP leaders to see exclusively how autonomous agents are reshaping their enterprise workflows, from real-time decision-making to end-to-end automation.
Secure your spot now – Space is limited: https://bit.ly/3guplf
Seven-digit monthly revenue spikes show that enterprises are making big bets on AI automation
According to Mlejnsky, the funds reflect explosive revenue growth, with E2B adding “seven numbers” in its new business over the past month. The company has handled hundreds of millions of sandbox sessions since October, showing the size of companies deploying AI agents.
E2B’s customer roster uses advanced data analytics capabilities for Pro users using AI Innovation: Search Engine Perplexity, using E2B to implement the functionality in just one week. AI chip company GROQ relies on E2B for secure code execution for its combined AI systems. The workflow automation platform Lindy Integrated E2B has enabled the execution of custom Python and JavaScript within user workflows.
Startup technology has also become an important infrastructure for AI research. The leading AI model repository, Hugging Face, uses E2B to safely execute code during reinforcement learning experiments to replicate advanced models such as DeepSeek-R1. Meanwhile, UC Berkeley’s Lmarena Platform has launched over 230,000 E2B sandboxes to evaluate the web development capabilities of large-scale language models.
Firecracker MicroVMS Solve dangerous code problems that plague AI development
E2B’s co-innovation comes from the use of Firecracker MicroVMS, a lightweight virtual machine originally developed by Amazon Web Services. This addresses the fundamental security challenges. AI agents often need to run untrusted code that can damage the system, or untrusted code that can potentially access sensitive data.
“When talking to customers and special companies, their biggest decision is almost always build and buy,” Mlejnsky explained in an interview. “With a build-to-purchase solution, it all comes down to whether you want to spend the next six to 12 months building an infrastructure team of 5-10 people that costs at least $500,000.
The platform supports multiple programming languages such as Python, JavaScript, and C++, allowing you to spin up new computing environments in around 150 milliseconds.
Enterprise customers particularly value E2B’s open source approach and deployment flexibility. Companies can self-host the entire platform for free or deploy it within their own virtual private cloud (VPC) to maintain data sovereignty.
Shifting to swap perfect timing AI workers as Microsoft’s layoffs
Funding comes at a critical moment for AI agent technology. Recent advances in large-scale language models have enabled AI agents to handle complex, real-world tasks. Microsoft recently fired thousands of employees, hoping that AI agents would previously do human-only jobs, Mlejnsky noted in our interview.
However, infrastructure restrictions limit the adoption of AI agents. Industry data suggests that less than 30% of AI agents will succeed in their production deployment due to the security, scalability and reliability challenges that E2B platforms aim to solve.
“We’re building the next cloud,” Mlejnsky outlined the company’s ambitious vision. “The current world runs on Cloud 2.0, made for humans. We can build an open source cloud for AI agents, and run it autonomously and securely.”
Market opportunities seem to be considerable. While the code generation assistant already produces at least 25% of the world’s software code, JPMorgan Chase saves 360,000 hours a year through its document processing agent. Enterprise leaders expect to use AI agents to automate 15%-50% of manual tasks, creating a huge demand to support their infrastructure.
Open Source Strategy creates defensive moats for high-tech giants like Amazon and Google
E2B faces potential competition with cloud giants such as Amazon, Google, Microsoft, and theoretically replicates similar features. However, the company has built competitive advantages through an open source approach, focusing on AI-specific use cases.
Mlejnsky explained that “we really don’t care” about the underlying virtualization technology, and that E2B focuses on creating open standards for how AI agents interact with computing resources. “We actually partner with many of these cloud providers because many companies want to actually deploy E2B within their AWS accounts.”
The company’s open source sandbox protocol is the de facto standard, with hundreds of millions of computational instances showing their actual effectiveness. This network effect makes it difficult for competitors to move E2B when companies are standardized on the platform.
Alternative solutions such as Docker containers are technically possible, but lack the security isolation and performance characteristics required for the deployment of production AI agents. According to Mlejnsky, building similar features in-house typically requires 5-10 infrastructure engineers and an annual cost of at least $500,000.
Enterprise features such as 24-hour sessions and 20,000 simultaneous sandboxes will drive Fortune 100 adoption
E2B’s enterprise success comes from features designed specifically for large-scale AI deployments. The platform can expand from 100 simultaneous sandboxes in the free tier to 20,000 simultaneous environments for enterprise customers, with each sandbox running up to 24 hours a day.
Advanced enterprise features include comprehensive logging and monitoring, network security controls, and secret management – essential features for Fortune 100 compliance requirements. The platform integrates with existing enterprise infrastructure to provide the required control of security teams.
“We have a very strong inbound,” Mlejnsky said, explaining the sales process. “If you work on 87%, you’ll be back for 13%.” Customer dissent is usually focused on security and privacy management rather than basic technology concerns, indicating the broad market acceptance of core value proposals.
Insight Partners’ $21 million BET validates AI infrastructure as the next major software category
Insight Partners’ investments reflect an increasingly trusted investor in AI infrastructure companies. Global software investors, who manage more than $90 billion in regulatory assets, have invested in over 800 companies worldwide, with 55 portfolio companies achieving early official services.
Praveen Akkiraju, Managing Director, Insight Partners, said: “This rapid growth and corporate adoption can be difficult to achieve. We believe that E2B’s open source sandbox standard will serve as the basis for safe and scalable AI adoption in Fortune 100 and above.”
The investment will fund the expansion of San Francisco’s E2B engineering and market teams, develop additional platform capabilities, and support a growing customer base. The company plans to enhance its open source sandbox protocol as a universal standard while developing enterprise-grade modules such as Secrets Vault and monitoring tools.
Infrastructure Play that lets you define the next chapter of Enterprise AI
The E2B trajectory reveals fundamental changes in how companies approach AI deployment. While much attention has been drawn to large-scale language models and AI applications, the company’s rapid adoption among Fortune 100 companies shows that specialized infrastructure has become a critical bottleneck.
Startup success highlights broader trends. As AI agents move from experimental tools to mission-critical systems, the underlying infrastructure requirements are much more similar to those of traditional enterprise software than consumer AI applications. Not only model performance, but security, compliance, and scalability determine which AI initiatives will be successful at scale.
For enterprise technology leaders, the emergence of E2B as a critical infrastructure suggests that AI transformation strategies must explain more than just model selection and application development. Companies that successfully extend AI agents will become companies that invest early in specialized infrastructure layers that enable autonomous AI operations.
In an age where AI agents are poised to handle the ever-growing share of knowledge work, platforms that keep those agents running safely can prove to be more valuable than the agents themselves.
Source link
