The Breakdown.
X-ray of a single vendor or product. Breakdowns drill deep into investor and financial signals, intellectual property filings, technical documentation, and customer feedback. Useful for separating real capabilities from hype, flagging vulnerabilities, and understanding risk.
See the facts without the spin.
A breakdown is built for a buy, partner or compete decision on a single vendor. Start with the call, the factual spine, and who actually builds the thing.
Bottom line up front
The call on the vendor first, fit, standing, and the one thing that should change your decision, before the supporting detail.
↳ The decision-relevant answer in fifteen seconds.
Northwind Defense AI is a credible agentic-SOC contender with real autonomy in narrow domains, but revenue concentration and a single-cloud dependency make it likely fragile under a hyperscaler price move.
Company & product
The factual spine, corporate identifiers, funding, headcount, product surface and deployment model, resolved and sourced, not lifted from the website.
↳ The vendor's real shape, on one card.
Who actually builds it.
Leadership and the teams beneath them, resolved from filings and hiring data, so you can see where the weight sits and who owns the claims. Hover a role.
Founder-led, engineering-heavy, three of four functions report into a co-founder. Hover a role to see the team behind it.
Patents & IP filings.
Filings are a signal you can read before a product ships. We resolve the complexity and map each claim to reality.
Hover the assembly to read the claim each component carries, and the one place the estate is conspicuously bare.
Financial & investor read.
Filings, hiring, and investor signals to help you read against the narrative. Learn where momentum is real and where the curve is bending.
Est. ARR. ARR has roughly 5×'d in eight quarters, a durable top line.
Hype vs. reality.
Every headline claim, held against the evidence. The bar shows how much of each promise the record actually supports and where the gap is pure marketing.
Autonomy holds in two narrow playbooks; everything else still gates on a human reviewer. · 45% claim gap
Benchmarked at 0.7s median across a 2M-edge graph, the claim substantiates. · 5% claim gap
Cloud-only for the newest detections; the on-prem build trails by about two releases. · 40% claim gap
No contract we reviewed actually commits to this figure, it lives in the marketing only. · 85% claim gap
The ecosystem.
The competitors, partners, and dependencies the vendor operates inside. Hover any node to trace the relationship and see which dependencies have no failover.
Hover any node to trace the relationship. Dashed red edges are dependencies with no failover.
Risk, receipts,
and the grade.
The parts that don't make the pitch deck, what could break, where every claim traces to, and the adversarial pass behind the verdict.
Risk flags & disclosures
Every material risk we surfaced, concentration, dependency, governance, litigation, graded by severity, with the disclosure (or its absence) noted.
↳ The risks that don't make the pitch deck.
Evidence register
Every source behind every figure and claim, graded on the NATO Admiralty code and hyperlinked to origin. Audit any line back to where it came from.
↳ Nothing asserted that you can't verify yourself.
Graded & adversarially checked
Every source carries an Admiralty grade and every estimate uses calibrated ICD 203 language. Before delivery, an independent red-team pass tries to break the verdict.
↳ The grade is the product. Always included.
Got burning questions?
Get the Breakdown.
$199 per breakdown, ~48-hour turnaround, free 90-second preview first. No contract.