← Home

Dust Closes $65M Series B for Enterprise AI Agents. The Multiple Is the Story.

Dust raised $65M Series B for enterprise AI workplace agents on a busy May 18 funding day. The undisclosed valuation is the signal worth reading.

Dust closed a $65 million Series B on May 18, with the funding earmarked to expand its enterprise AI workplace agents platform. The round joined a notable May 18 funding day that also included Photonic's $75M Series C for quantum networking infrastructure and Destinus's $50M Series B for hypersonic defense systems. Dust did not disclose a post-money valuation.

The interesting cap-table read is the absence of a public valuation. Enterprise-AI-agent companies that disclose at this round size are typically pricing at 30-50x annualized revenue, depending on growth rate and customer concentration. Dust holding back the number signals two possibilities. Either the company priced inside that band and is downplaying the multiple to avoid the comp-set pressure that comes with public pricing. Or the round was structured with significant secondary, anti-dilution, or other protective terms that make a clean post-money number hard to publish. Both readings are consistent with a market where enterprise-AI-agent valuations have run ahead of demonstrable revenue. The category context matters. Anthropic's $1.5B JV with Goldman, Blackstone, and Hellman & Friedman, and Anthropic's new partnership with PwC, are absorbing the high-end enterprise-services demand. Independent agent platforms like Dust need to demonstrate why their product wins in the procurement room when the customer's existing AI vendor is offering the same workflow as a built-in. The $65M needs to fund go-to-market that wins those decisions.

For Series B-stage AI agent investors: the question is whether Dust priced above or below the published Anthropic-affiliated alternatives. The round size is less informative than the pricing signal. For enterprise buyers: include independent agent platforms in your bake-off through Q3. The discount to first-party labs is real but narrowing.