A novelty search looks for publicly available information that may be relevant to whether an invention is new before a filing decision. It starts with the technical features that matter, then looks beyond matching titles or abstracts to patents, published applications, and, where relevant, non-patent literature. In Eureka, the agent turns an invention description into search strategies, ranks potentially relevant references, and maps passages back to the disclosed features. The result helps a team prepare for a filing discussion; it does not determine patentability, replace examination by a patent office, or substitute for advice from qualified patent counsel.

A novelty search asks whether the technical idea appears to have been disclosed before, which is useful when evaluating an invention before filing. A freedom-to-operate, or FTO, search asks a different business question: whether a planned product or process could fall within claims that are enforceable in specific target markets. FTO work therefore needs product features, intended jurisdictions, claim analysis, and legal-status context. A positive novelty result does not establish freedom to operate, and owning a patent does not automatically give a company the right to commercialize a product.

No. A granted patent generally gives its owner the right to exclude others from making, using, selling, or importing the claimed invention in the relevant jurisdiction. It does not guarantee that the owner can use or sell a product without encountering another party's rights. A product can include an improvement that is patentable while still relying on technology covered by an earlier, broader patent. Before launch, teams usually need to assess the product as it will actually be made, used, and sold in each target market, then seek qualified legal advice on the relevant risks.

An effective FTO search begins with a defined commercial scenario: the product or process, the features that may matter, the countries or regions involved, and the expected timing of launch. The search should identify potentially relevant patent rights and published applications, then examine the claims and legal-status context rather than relying only on titles, abstracts, or similarity scores. Eureka helps structure this evidence into claim-level comparisons for review. The output is designed to help legal and product teams prioritize questions, but it is not an infringement opinion or a guarantee that no enforceable rights exist.

AI can reduce the time required to structure a search, explore terminology, retrieve potentially relevant documents, and organize evidence for review. It should not replace the professional judgment needed to interpret claim scope, assess legal status, account for jurisdiction-specific rules, or provide legal advice. Eureka is designed for a human-in-the-loop workflow: the agent prepares search strategies and reviewable evidence, while IP professionals decide what is relevant, what requires deeper analysis, and what action to take. This division of work helps teams move faster without treating an AI response as a legal conclusion.

Design-risk review should start with the actual appearance of the product, including the visual features most likely to shape overall impression. Teams should consider the relevant design registrations and applications in the markets where the product will be manufactured, offered, or sold, then review visual similarity in context. Eureka converts product imagery into patent-standard line drawings, retrieves visually similar design registrations, and organizes the visual evidence for review. The output can help surface potential issues before tooling or release, but the final assessment of infringement risk requires qualified legal analysis in the relevant jurisdiction.

Patent searching often begins before an invention, product plan, or design has been made public. Eureka is designed for that context: queries, saved patents, uploads, and analysis outputs are not used to train its models; data is encrypted at rest and isolated within enterprise-grade infrastructure. Users control what they upload and when they delete it, while enterprise teams can use controls such as SSO, audit logs, and role-based permissions. These safeguards support confidential research workflows, but each organization should still apply its own information-governance, access-control, and legal-review policies.