Trigger: On October 5, 2026, Cyberhaven posted a US-remote Director, Product Design role that partners with the CPO on data-security experiences for human analysts and AI agents. Confidence is high on the role and moderate on the date.
Window
Today is day 25, inside the Watch band (days 22–42). Unless something new happens, the posting drops off our priority list on November 17 (day 43). That is our review date, not Cyberhaven's deadline; the company hasn't published a closing date.
Redpoint's portfolio board shows October 5, the same day Remote Rocketship first picked up the role. UknowUX, an independent listing service, shows October 20. Board dates record when an aggregator found the role, which can be well after Cyberhaven started recruiting, and Cyberhaven's own Ashby timestamp isn't retrievable, so the date stays at moderate. Title and Ashby link match across every listing, so there's no sign of a repost. Load the application form yourself before you spend an evening on it.
Manish Kalia was named CPO on August 19, in the same release that named a new President. He's about 72 days in, which falls in our 61–90 day staleness band for executive hires. Treat the posting as the live signal; the CPO hire is already stale as a trigger on its own.
Company snapshot
Cyberhaven sells data loss prevention (blocking sensitive data from leaving), insider-risk detection, and DSPM (finding and classifying sensitive data wherever it sits). All three run on lineage: the system records where each piece of data came from, along with every copy, split and merge since. Most DLP tools inspect content only at the moment it leaves, while Cyberhaven can judge a file by its history. AI handles detection and investigation and writes up each case in plain language. Flow, announced July 28 for "the coming quarter," merges lineage, identity and behavior across human and agent workflows. Buyers compare Cyberhaven with Microsoft, Proofpoint and Netskope, and with Cyera for DSPM.
What this trigger reveals
Cyberhaven named a new CPO in August. Its former Head of Design now runs a security-design consultancy. Seven weeks after the CPO announcement, this req went up. Read together, that looks like a CPO rebuilding design around Flow, including the work of moving existing customers onto it. Moderate confidence. The read rests on timing, and the design head's departure is undated, so this could just as easily be a backfill as a new seat.
Fit is harder. The JD asks for someone who has "worked in cybersecurity, or designed for security operations, insider risk, or data protection teams." As written, you don't meet it. The gap is in who the user is. At TinyFish, operators audited their own agents, and the trace helped them trust a tool they already owned. A Cyberhaven analyst reads traces about employees, or about unapproved agents that may be the threat, and works through heavy alert volume where both mistakes cost something real: a missed exfiltration, or an innocent employee blocked. Your governance record carries over to the agent-security half of the job. Traditional insider risk is your thinnest ground.
Rubric scores
| Dimension | Score | Reasoning |
|---|---|---|
| AI centrality | 2 | AI runs detection and explanation. Lineage capture and blocking don't need it. |
| Stage/equity | 3 | Pre-IPO Series D at $1B, April 2025. Grant terms are undisclosed. |
| Design influence ceiling | 2 | The JD promises an executive-planning seat. The team is small and the reporting line is unstated. |
| Trajectory | 3 | Self-reported triple-digit growth for the year ending January 2026. |
| Company total | 10 | 2 + 3 + 2 + 3 |
| Total comp | 3 | Posted $250K–275K plus equity and bonus. Pay is location-adjusted, so this sits at the bottom edge of 3. |
| Scope expandability | 2 | Top design seat with a plausible VP path. Small team. |
| Craft depth | 3 | "Hands-on leadership role," partnered directly with the CPO. |
| AI exposure quality | 2 | You'd design the AI layer, at a company that isn't AI-native at its core. |
| Portfolio value | 3 | Humans overseeing agents that police other agents is rare work. Security NDAs may limit what you can show. |
| Role total | 13 | 3 + 2 + 3 + 2 + 3 |
Company 10, role 13: Act. Drop trajectory and comp one point each and the totals fall to 9 and 12, still Act, right at the threshold.
Landscape context
Ground Truth has no data-security coverage yet; the category sits outside our eight verticals. From outside sources: Cyera raised $600M at $12B in June and another $400M in September, and Netskope IPO'd at $7.26B. Cyberhaven has gone nearly 19 months since its last announced priced round. A new round could reprice your equity upward, but until one arrives, Cyberhaven has less money than its rivals for the race to secure agents.
Verdict
The score says Act and the clock says Watch. The security-experience requirement decides whether either matters for your candidacy.
Assume that line is required. Nothing in the posting waives it, and a nearby line about expert users in IT operations, observability and data infrastructure sits next to it without replacing it. A high score rates the opportunity but can't clear a requirement on your side. Agent security is your one credible route to that line; if it can't carry the requirement, the Act score does nothing for you.
Keep the tracker slot through November 17. When I overrode Peregrine's Act in an earlier piece, it was because ownership of the AI's behavior was unresolved. This JD settles that question by giving design "human-in-the-loop decisions, explainability," so the score stands. Few Director roles put human oversight of agents at the center of the product.
Warning signs
- You'd migrate customers to a platform without confirmed general availability. Flow was announced July 28 with early access, and no GA confirmation has surfaced since.
- Leadership churn. Cyberhaven got a new CEO in May 2025 and a new CPO and President in August 2026, and the design head is gone. One May 2026 Glassdoor review alleges layoffs, which the company denies. A single review is weak evidence; watch for the claim to repeat.
- Flow's general availability: The migration half of this mandate depends on whether Cyberhaven confirms GA for Flow, which was announced July 28 for "the coming quarter" and has offered only early access so far.
- Read-only agent oversight: Cyberhaven describes its Claude Enterprise connector as read-only, meaning it inspects chats and files but cannot control Claude, which marks where the design problem shifts from observation to enforcement.
- Backfill or new seat: The former Head of Design's consulting practice confirms she has moved on but gives no departure date, so a dated exit would settle whether this req replaces her or adds a new leadership layer.
- Cyera's capital lead: Cyera added a $400M Series G extension from Goldman Sachs in September while pushing into AI-agent security, the same ground Cyberhaven is contesting on nearly 19-month-old Series D money.
