Investor information
This page is deliberately brief and deliberately unpromotional. Financial detail, the territory model and the operating plan sit in a data room shared under NDA on request.
The company
Densery builds governed AI agents that complete document-driven work inside regulated institutions — know-your-customer and loan file review, claims and medical records, underwriting evidence, engineering and compliance records — and write the result into the customer's system of record with the evidence a supervisor will accept. It deploys on the customer's own hardware where data residency is a legal obligation.
The business is not pre-revenue and not a pilot. It has six named production deployments across banking, insurance, auto finance, manufacturing and digital infrastructure, spanning Vietnam, Japan and Indonesia, with demonstrated return in three regulated verticals.
The plan is to take a business that already works, in the hardest deployment conditions available, into the largest market that has the same constraints.
The US thesis
47% of banking and insurance firms already run an AI agent in production, against 31% of enterprises overall. But 88% of agent pilots never reach production and 95% of generative-AI pilots show no measurable P&L impact. The budget and the mandate exist; the delivery does not.
Leaders cite an evaluation gap (64%), governance friction (57%) and model reliability (51%). Densery's architecture was built against those three constraints in markets where they were legal requirements rather than preferences.
The US plan is built bottom-up from 145 scored, named accounts across six segments, filtered on document volume, regulatory burden, deal-size fit, residency obligation and reference adjacency — not from an analyst headline. The resulting market is deliberately smaller than the top-down number, and it is the one a seller can actually work.
What we think is hard
Any investor doing real work will find these. We would rather they arrived from us.
Revenue, growth rate, retention, burn, valuation, cap table, the named account list and the operating plan. Those are in the data room, and they are shared with investors who have been through an introductory conversation — not because they are unflattering, but because a public page is the wrong place for them and our customers read this site too.
We respond to every legitimate request, usually within two business days.
Materials are shared under NDA. We do not add anyone to a mailing list from this form.
Before the financials, ask for a reference call. Six institutions run this in production and any of them will tell you what it was actually like.