How AI Briefings work

An AI Briefing is a written research note generated for a specific ticker, combining this platform's own valuation numbers with a search for recent, relevant news — not a generic summary from a general-purpose chatbot.

What grounds it

Every briefing starts from the same Sticker Price, Payback Price, verdict, and confidence score already computed for that ticker elsewhere on the platform — these numbers are treated as ground truth and are never re-derived or second-guessed by the AI narrative. The model's job is to add context around them: recent news, notable filings, insider activity, and anything else relevant to whether the current price makes sense in light of the valuation — not to produce a competing set of numbers.

Why it searches the web

The model runs a small number of targeted searches while writing a briefing, specifically to pull in information more recent than its own training — a recent earnings call, a management change, breaking news. This is what lets a briefing say something concrete about this week, not just a static description of the business.

Why briefings are cached

A briefing is generated once per ticker and reused for 30 days, shared across every user who requests that ticker in that window — re-requesting an already-generated briefing is instant and free. Generating a genuinely new one involves a real AI model call plus several web searches, which is why it isn't unlimited.

The monthly limit

Each account gets 3 new briefing generations per month. This only counts a request that actually triggers new generation — reading a briefing someone else (or you) already generated recently doesn't use any of your allowance.