Sequoia just doubled down on agent identity. The Series A is small. The category question is not.
On September 9 TechCrunch reported that Cymphony emerged with $30 million in funding to help enterprises govern a growing AI workforce. The package includes a $25 million Series A co-led by Sequoia Capital and SMBC Fin Atlas Beyond Fund, valuing the New York– and Tel Aviv–based startup at more than $100 million after investment. A previously undisclosed Sequoia seed sits underneath. Those are the only valuation and round figures, per TechCrunch’s exclusive.
The product thesis is blunt. Enterprise security was designed for human employees. Agents now join the workforce as independent entities with multi-system access and machine-speed data handling, often outside the same access and identity controls humans get. Cymphony’s answer is a “workforce graph” that unifies identity, data, and activity signals across employees, AI agents, and other nonhuman identities. Customer anecdotes in the piece — not a market size: at one U.S. public company, about 85,000 files had become accessible to AI tools and agents; Cymphony says it helped close the exposure and verified none were accessed through those AI systems. Separately, an external collaborator installed an unsanctioned Claude instance and used existing access to scan thousands of sensitive files. Remediation includes AI agents that investigate, prioritize, and automate some permission fixes — plus an optional managed service.
Traction claims from the startup to TechCrunch: product shipped; a double-digit number of enterprise customers; seven-figure ARR within the first year of sales; named logos include KKR, Syngenta, Cass Information Systems, and Athennian. Headcount about 30 across Tel Aviv and New York; customers mostly North America, with early EMEA demand. Sequoia partner Bogomil Balkansky framed the seed as a Talpiot-founder bet before product-market fit, and the Series A as proof after product, logos, and expansion inside existing accounts — Sequoia has also used the product internally. Competitive set named in the article: Microsoft, Okta, CyberArk, Wiz, Varonis. Balkansky’s line worth keeping: nobody is ripping out Okta tomorrow; Cymphony is largely an additional layer today, with possible later displacement of some point solutions such as DLP.
POV: this is the same tape as Cyera’s Oasis NHI stitch and every “agent security is a feature” platform slide — just earlier and thinner. A >$100M post-money on $25M Series A with seven-figure ARR is a believable seed-to-A mark for 2026 cyber. It is not proof agent security is a standalone category. Sequoia’s Balkansky says if companies are not spending on agent security, he does not know what else they will spend on in the next five to ten years. That is a thesis, not a TAM. Underwrite the graph (identity + data + activity) versus platform attach. Question whether Okta/CyberArk/Wiz absorb the layer before the startup owns the budget line. Source: TechCrunch. ARR dollars beyond “seven figures”: Undisclosed.
Underwrite sheet — sourced only: $30M total funding package; $25M Series A co-led Sequoia + SMBC Fin Atlas Beyond; >$100M post-money; prior undisclosed Sequoia seed; workforce graph (identity/data/activity); ~85k files anecdote; unsanctioned Claude collaborator anecdote; double-digit enterprise customers; seven-figure ARR in year one of sales; logos KKR/Syngenta/Cass/Athennian; ~30 employees; N.A. heavy; competitors Microsoft/Okta/CyberArk/Wiz/Varonis; complementary-to-Okta framing (TechCrunch, Sep 9, 2026). Category ≠ feature is the fight. The $100M mark is not the product.