Daytona reached $1M forward revenue run rate in under three months. Six weeks later, it doubled.

The speed reflects enterprise urgency around a fundamental infrastructure gap: current cloud platforms were built for production workloads—stateless, immutable systems optimized to run the same code the same way every time. AI agents need the opposite: persistent, stateful environments where they can experiment, branch execution paths, and recover from failures at massive scale.

The Infrastructure Mismatch

Enterprise adoption of AI agents is hitting a hard infrastructure bottleneck. While agents can reason through complex workflows and make decisions, they break when deployment infrastructure treats every workload like a web server.

EnFi raised $15 million in Series A funding to deploy AI agents that function as virtual credit professionals, addressing a fundamental scaling crisis where commercial lending demand is growing faster than the human talent pool can accommodate.

The Boston-based startup represents a shift from AI-assisted tools to autonomous agent workforces, with customers deploying targeted agents across deal screening, documentation, and portfolio monitoring within 60 to 90 days. This timeline reflects a new deployment velocity for financial services AI, where traditional implementations often require 12-18 months.

79% of enterprises now deploy AI agents, but 97% lack proper security controls—a gap that’s created a new category of infrastructure risk as organizations race to automate everything from database queries to payment processing.

Virtue AI just launched AgentSuite, the industry’s first multi-layer security and compliance platform purpose-built for enterprise AI agents. While traditional security tools were designed for predictable application workflows, AgentSuite addresses the dynamic, multi-tool nature of autonomous agents that can trigger complex actions across enterprise systems in real time.

Ricursive Intelligence has raised $300 million in Series A funding at a $4 billion valuation, just two months after the company’s formal launch. The round was led by Lightspeed Venture Partners, with participation from DST Global, Nvidia’s NVentures, Felicis Ventures, Sequoia Capital, and others.

The funding milestone reflects investor confidence in what the company calls “recursive AI-hardware co-evolution”—AI systems that design and continuously improve the chips that power them. For an industry where chip design cycles have become a critical bottleneck to AI advancement, Ricursive’s platform promises to collapse months-long design processes into hours while achieving superhuman layout optimization.

Waabi secured $1 billion in total funding this week—including a $750 million Series C round co-led by Khosla Ventures and G2 Venture Partners, plus milestone-based investment tied to its robotaxi partnership with Uber—marking one of the largest funding rounds in Canadian tech history.

The Toronto-based company represents a critical infrastructure shift: while most AI agent development has focused on digital environments, Waabi’s Physical AI platform bridges autonomous agents into real-world deployment at enterprise scale. This funding validates Physical AI as the next major infrastructure category, where agents must navigate safety-critical environments rather than just process digital workflows.

Poetiq’s $45.8 million seed round, announced Wednesday, comes with proof that six former Google DeepMind researchers have cracked a fundamental barrier in AI reasoning—achieving 75% accuracy on the notoriously difficult ARC-AGI-2 benchmark, a 16-percentage-point leap beyond previous state-of-the-art.

The achievement matters now because enterprises are drowning in a $30-40 billion AI investment crisis. An MIT study published in August 2025 found that 95% of organizations are “getting zero return” on their GenAI deployments, with most failures traced to LLMs’ inability to handle real-world reasoning tasks that require more than pattern matching.

Limy emerged from stealth today with $10 million in funding to address what its founders call the “agentic web” infrastructure gap—the disconnect between traditional user-centric analytics and the AI agent behaviors that increasingly drive online discovery and commerce.

The funding round, led by Flybridge with participation from a16z speedrun, signals investor recognition of a fundamental shift: As AI agents become the primary interface for information discovery, brands need entirely new infrastructure to understand and optimize for agent behavior rather than human clicks.

Fiddler AI closed a $30M Series C led by RPS Ventures, bringing total funding to $100M as enterprises confront the reality that AI agents require fundamentally different infrastructure than traditional software—not just monitoring, but active control.

The governance gap has real consequences. Agent failures can trigger regulatory fines, legal exposure, and brand damage, driving rapid adoption among Fortune 500 companies where AI oversight isn’t optional. Fiddler achieved 4x revenue growth over 18 months and earned AWS Pattern Partners status as organizations realize that deploying autonomous agents without proper controls creates existential risk.

LiveKit raised $100 million Series C at a $1 billion valuation, led by Index Ventures with Salesforce Ventures, Hanabi Capital, Altimeter, and Redpoint Ventures participating. The open-source-born infrastructure provider powers the real-time voice capabilities behind OpenAI’s ChatGPT voice mode, xAI’s Grok Voice Agent API for Tesla vehicles, and hundreds of enterprise voice AI deployments.

The funding addresses a fundamental infrastructure bottleneck: building production-ready voice AI requires orchestrating speech-to-text, language models, text-to-speech, and real-time communication protocols—a complex technical challenge that most companies struggle to solve in-house. LiveKit abstracts this complexity into a unified platform, turning months of infrastructure development into minutes of deployment.

Inferact launched Wednesday with $150 million in seed funding at an $800 million valuation to commercialize vLLM, the open-source inference engine that reduces AI deployment costs by up to 70%. The round, co-led by Andreessen Horowitz and Lightspeed Venture Partners, represents one of the largest seed valuations ever and signals a fundamental shift in AI industry priorities from model training to deployment optimization.

The infrastructure bottleneck is real. Organizations deploying AI applications are discovering that inference costs—the expense of running trained models to generate outputs—often exceed training expenses over a product’s lifetime. Companies like Stripe report 70% cost reductions using vLLM, while the technology enables significantly faster processing and higher hardware utilization across the AI stack.