Six posts going up Tuesday morning. Engage on all of them. If you want to make your own post, use the company-wide draft and personalize it. Research and release URLs go in the first comment, never the body.
Posting plan for Tuesday May 26
Lior posts first at the time the wire goes out. Yonatan, Israel, Brian, and Nate post within 90 minutes after, each angled to their function. The 7AI company page posts the company-wide version simultaneously with Lior. Engagement on every post in the first two hours is what creates the compounding signal. Like every post, leave a comment that adds something beyond the post itself, repost with your own short framing.
LD
Lior Div / CEO & Co-Founder
Posts first / compounding advantage angle
A year ago we said AI agents could fundamentally change how security operations work.
After 7 million investigations in production at Fortune 500 scale, here is what we have actually learned:
The compounding advantage is bigger than we anticipated. Every investigation makes the next one faster, more accurate, more contextual. Agents that have worked across 7 million investigations understand environments, attacker patterns, and signal interpretation in ways that cannot be replicated by reading a brief or running a pilot.
That gap does not shrink. It grows.
Today we are sharing what a year at enterprise scale looks like, announcing PLAID ELITE as broadly available, and committing 100 new AI jobs to Boston in 2026.
7ai.com/momentum (link in comments)
The interesting technical detail in our Momentum announcement is not the 7 million investigations number itself. It is what 7 million investigations does to the system that processes them.
Every investigation enriches the system's understanding of attacker behavior across environments. Every signal interpretation accumulates into context that the next investigation inherits. Every false positive surfaced and resolved is a data point that sharpens future detection.
This is not training in the traditional sense. The agents are not waiting for a quarterly model update. The compounding happens in production, continuously, across the entire customer base.
The closest analog is how a senior detection engineer gets better over a decade. Except the platform is doing it across millions of investigations at once, in weeks.
That is what production-scale agentic security actually means. Not "we have AI." A system whose floor on day 365 is higher than its ceiling on day 30.
When I joined 7AI as CISO in March, the thing that drew me in was the managed offering.
I have spent years running security operations at scale. The structural problem with traditional managed services is not the analyst quality. It is the delivery model. Coverage and response time are bounded by how many humans are on shift. Cost scales linearly with capability. The unit economics break the moment a customer needs both depth and breadth.
PLAID ELITE solves that, and as of today it is broadly available.
Continuous autonomous investigation by AI agents handles the work that scales. 7AI security engineers add the judgment, the context, and the response orchestration that humans should be doing. Performance improves with investigation volume rather than scaling with headcount.
That is a fundamentally different service model than anything that came before it. And for the first time, organizations that want agentic SOC outcomes do not have to build an internal operations team to get them.
The Momentum release is out and the headline numbers are the headline numbers.
But the ones I am proud of, because of what they say about the partners doing this work with us, are these:
6.5x channel pipeline expansion in three quarters.
45% of total pipeline sourced through the channel.
7x growth in partner-registered wins quarter over quarter.
We made a call from the beginning that 7AI would be channel-first. No direct competition with our partners. No second-guessing where deals were sourced. Real co-sell motions where the partner is the trusted relationship and we are the technology that makes their advisory work compound.
A year in, the result is that our fastest-growing pipeline category is partner-led, and our newest service offering, PLAID ELITE, is structurally designed to be wrapped inside a partner's managed practice.
GuidePoint, Myriad360, DXC AdvisoryX, and the rest of the partner community, thank you. The category is moving because of how you are showing up to the customers we serve together.
Boston has been quietly building the next generation of AI security in plain sight.
Today we are adding to that story:
100 new AI jobs at 7AI's Boston HQ in 2026, across AI security engineering, product, and go-to-market.
A founding membership in the Massachusetts AI Coalition, alongside companies like WHOOP, anchoring Massachusetts as a national AI hub.
Hosting the Boston Tech Week cybersecurity kickoff at F1 Arcade Seaport on May 28 with a16z, GuidePoint, Astrix, Cogent, Upwind, and AWS. Julian Edelman keynoting. A panel on AI security in production with Benjamin Dulieu from Duck Creek, Joel Miller from DraftKings, and Debashis Das from AWS.
We chose to build 7AI in-person in Boston for a reason. The work this category requires (research, threat hunting, product engineering, customer trust) is collaborative work that does not replicate over video.
If you are in Boston and want to be part of this, we are hiring across the board. If you can be at F1 Arcade Thursday night, come say hello.
A year ago we were running a few thousand investigations a day. Today we are past seven million in production.
The engineering reality behind that number is not glamorous:
Token economics that make per-investigation cost predictable enough to write into a contract.
Plan execution that keeps an agent honest about what it actually ran versus what it claimed to have run.
Replay infrastructure that lets a human reconstruct any decision the system made, on demand.
Eval harnesses that catch model regressions before they reach a customer environment.
None of that gets a press release. All of it is what makes the press releases possible.
We are hiring AI security engineers in Boston in 2026. If this sounds like the kind of problem you want to spend your time on, come build with us.
Today we are sharing one year of what agentic security looks like at production scale:
More than 7 million investigations completed by AI agents in customer environments.
3x customer growth quarter over quarter.
6.5x channel pipeline expansion in three quarters.
95 to 99% false positive reduction reported by customers within 90 days.
PLAID ELITE, our fully managed AI-native security operations service, is broadly available.
100 new AI jobs at our Boston HQ in 2026.
The agentic security category is no longer about whether AI agents can do the work of a SOC. It is about what compounds when they do.
Read the full announcement at 7ai.com/momentum (link in comments).
Trusted by DXC Technology, BigID, Duck Creek Technologies, Blackstone, OneSpan, Cole Scott & Kissane, and a growing list of customers.
If you want to engage on an exec post with a comment that adds signal beyond a like, pick one of these and personalize the bracketed pieces.
The compounding number that sticks with me is the 6.5x channel pipeline expansion. Not because of the multiple, but because of what it implies: our partners are pulling 7AI into deals before we even know the customer exists. That is the strongest signal we are doing this right.
From the field, the part of today that I am hearing customers respond to is PLAID ELITE. Specifically, the fact that performance improves with investigation volume rather than scaling with headcount. That changes the conversation from "what do we buy" to "what do we get to stop doing."
The detail that stayed with me is "under 72 hours from contract to first autonomous investigation." That is not a marketing number. That is how fast our last [N] customer deployments actually went. Worth re-reading the release for the operational specifics.
Reshare with your own copy
If you want to reshare one of the exec posts with your own short framing, use one of these.
The category has moved. A year ago "agentic security" was a position paper. Today it is 7 million investigations in production, named Fortune 500 customers, and a fully managed service that is broadly available. Worth reading what that actually looks like.
The part of the Momentum story that does not get enough attention is the Boston anchor. 100 new AI jobs, founding membership in the Massachusetts AI Coalition, hosting Boston Tech Week's cybersecurity kickoff. AI security gets built in the cities that commit to it. Boston is showing up.