ASSET

The work behind each asset begins months before any operator conversation. Kalibar reads the sector across twelve dimensions of operating conditions — macroeconomic, geopolitical, regulatory, capital, resource, technological, supply-chain, competitive, demand-side, cultural, environmental, and tail-risk — and applies four lenses to each: what argues against a given direction, what argues for it, the single variable whose movement would flip the picture, and the order in which the signals will arrive. Sources are graded by reliability. Every prediction comes with the specific observations that would prove it wrong. The research and modeling produce two outputs: territories where genuinely new patentable subject matter is forming with economic power, and territories where filings would be technically valid but commercially obsolete because regulation, supply, or context will undercut deployment. The filing covers the first kind of territory and avoids the second.

Candidate acquirers are graded by how their enterprise trajectory overlaps a qualified forming territory, a “lighthouse zone.” They are scored across financial profile, sector position, IP-awareness, governance, and other criteria.

The inventive work begins after that. Every step is recorded in a bound evidentiary book, with each entry signed by the inventor and a witness. A prior-art library is assembled, with every reference analyzed and either incorporated or explicitly distinguished. Every AI session involved in the work is logged and used at enterprise level with no AI training permitted contractually. Formal attestations document Bayh-Dole compliance — confirming no government-funded research is implicated — and prior-employer clearance — confirming no employment agreement reaches the work.

The application then runs through numerous distinct attacks. Many look for legal grounds a court or examiner could use to invalidate the patent: anticipation by prior art, obviousness, ineligible subject matter, vague claim language, insufficient teaching of how to build the invention, missing demonstration that the inventor had the full invention at filing, unanchored functional claim language, and weak structural support for the action words in the claims. Some hardening steps target procedural and post-grant issues: whether the patent would survive a competitor's invalidation petition after grant, whether the assigned examiner has any disqualifying connections, and whether the strongest available claim type has been chosen. Various attacks assess design-around vulnerability: whether a competitor can design around the claim cheaply, and whether the patent has clear infringement detectability and claim boundaries. The process includes prediction of the USPTO art unit’s rejection patterns, factored in the construction.

Each asset includes collateral materials the acquirer may need downstream.

For licensing and enforcement, the package includes a unified claim chart set: a lineage chart showing how the claims developed and what they now cover, a licensing claim chart, a claim construction chart where one is warranted, and a litigated-logic folio showing how each claim element maps to successful claim-logic patterns that have prevailed in prior litigation. Each chart sits inside narrative front matter that defines the invention, identifies its function and structure, and quantifies its coverage relative to the existing art and inventive contributions of others.

For the acquisition conversation, the package includes a one-page deal sheet, a value dossier laying out the market and regulatory environment, the research and modeling report, a workaround estimation that quantifies what it would cost a competitor to design around the patent, a clean-title memo, and a purchase agreement.

The operator inherits the position itself, the construction record behind it, and every deliverable that supports its downstream use.

AI Infrastructure · Agentic Systems · Clinical Data & Instrumentation · Financial-Systems Automation