No, and the system prompt is explicit about this. Royalty rates, term sheets, and other commercial conditions are negotiated under NDA and the bot is instructed to refuse them politely and route to BD. The bot describes licensing models in general (exclusive, non-exclusive, field-of-use) and the kinds of structures available, which is enough for a prospective licensee to know whether to start the conversation.

Yes, as long as that data lives in ACF fields on the patent or portfolio post. A 'jurisdictions' multi-select with values like US, EU, JP, KR, CN lets the bot answer 'has this issued in China' with a real yes or no, and 'has this been granted in the EU' with the specific grant year if you store it. For pending applications, the bot can name the jurisdiction without invoking confidentiality on the application content itself.

OpenAI Files vector storage handles portfolios with hundreds of patents. Each patent abstract, claims summary, and metadata lives in the vector store and the bot retrieves the relevant subset for the conversation. For very large portfolios where the patent text itself is too dense for full indexing, a curated summary per family is usually the right granularity.

Yes, as long as the SEP status is captured on the patent record. Fields for SSO declarations (3GPP, ETSI, IEEE), FRAND commitments, and essentiality rank can live on each post. The bot quotes the public declarations and FRAND posture without entering negotiation territory. For SEP-heavy portfolios, this is often the first qualifying question from corporate development teams.

Yes. WPML or Polylang content is honored, and the system prompt can be tuned per language. A Japanese licensee asking about JP-issued patents gets a Japanese-language answer drawn from your Japanese content. For international licensing programs where the licensee pool spans regions, this preserves the relationship before BD ever picks up the conversation.

Yes. Display conditions show the bot only on public marketing pages and the portfolio overview. NDA-only client portals, confidential term sheet pages, or in-flight deal rooms can be excluded explicitly. For some firms, a separate bot scoped to those pages can handle logged-in client questions with a different system prompt that has been audited for confidential context.

Inside your WordPress database. Each log row stores the model used, token usage, the page URL, and the full conversation. For licensing firms with auditability or e-discovery obligations, in-database logs that you control directly are a meaningful advantage over hosted chatbot services. You decide retention, you decide who can access logs, and you can export them in standard formats if a litigation hold ever requires it.

No, and it should not. The system prompt is configured to politely refuse alleged infringement discussions and route them to legal counsel via your formal notice address. A chatbot is not the right channel for those conversations, both legally and operationally. The bot's job is to surface portfolio details for prospective licensees; legal correspondence stays on the formal channel.