Runlayer closed a $30 million Series A led by Felicis and Khosla Ventures to build governance infrastructure for enterprise AI agent deployments. Vinod Khosla pushed to acquire the entire round, according to Startup Fortune, bringing total funding to $42 million seven months after the company emerged from stealth with an $11 million seed from the same two firms.
The Product
Runlayer builds the layer underneath models and agent frameworks: a unified dashboard where enterprises can see which AI agents are running, what data they have accessed, what they cost, and whether they performed as intended. Employees connect agents from OpenAI, Anthropic, Salesforce, or internal systems through pre-approved channels with company data and guardrails already configured, per Startup Fortune. IT and security teams get one control room instead of scattered agent deployments across the business.
Founding Team
CEO Andrew Berman previously cofounded Nanit (AI baby monitors) and Vowel (AI video conferencing, acquired by Zapier in 2024), then served as Zapier’s director of AI. Co-founders Tal Peretz led machine learning for the Israeli Air Force, and Vitor Balocco was a staff AI engineer at Zapier. The Zapier connection matters: the company sits at the intersection of business tool integrations where small automations quietly become operational infrastructure.
Early Traction
Within four months of stealth launch, Runlayer signed Instacart, Gusto, dbt Labs, Opendoor, Homebase, and AngelList, according to Startup Fortune. The customer mix spans marketplace operations, payroll, developer tooling, real estate, and small business software, suggesting agent governance demand is not confined to a single vertical.
The company is also a founding member of the Anthropic-led AI Agent Interoperability Forum alongside OpenAI and Google, placing it near the standards conversation while those standards remain under debate.
Market Context
Gartner projects 40% of enterprise applications will include AI agents by the end of 2026, up from less than 5% in 2025, with agentic AI spending on pace to reach $201.9 billion this year, as cited by Startup Fortune. That proliferation creates the problem Runlayer targets: governance inside one platform (ServiceNow governing ServiceNow agents, Microsoft governing Microsoft agents) does not extend across the full agent estate. A vendor-neutral control layer has to sit above every model, every agent vendor, and every internal tool.
The Investor Signal
Khosla’s push to take the entire round follows the classic picks-and-shovels thesis in AI investing: back the infrastructure every application has to touch. A $30 million Series A with that level of investor conviction seven months after stealth suggests the governance gap between agent deployment speed and enterprise control capacity is widening faster than the platforms can close it internally.