Knowledge Base

AI features

How AI evaluation judges tenders against your filter, what Matched and Unmatched mean, the AI summary, the models, the assistant, and the limits of all of it.

  1. Article 1How AI tender evaluation works in Tender NavigatorHow Tender Navigator's AI evaluation judges each tender a filter picks up against the filter's prompt, what the model receives, what it returns (Matched or Unmatched with reasoning), and when it runs.Read article
  2. Article 2Understanding Matched and Unmatched resultsHow to read the Results page in Tender Navigator: the Matched and Unmatched verdicts, the reasoning, the filters and date range, the seven-day default window, and the actions on a result.Read article
  3. Article 3Why a tender was matched or unmatchedHow to find out why Tender Navigator's AI marked a tender as Matched or Unmatched: the reasoning on the result, the filter snapshot it was judged against, and how to change future verdicts.Read article
  4. Article 4The AI summary on tender noticesWhat the AI summary on a Tender Navigator tender notice is, how it is generated on first open, why it can appear a few seconds after the rest of the notice, and how it differs from AI evaluation.Read article
  5. Article 5Choosing the AI model for a filterThe LLM model setting on a Tender Navigator filter: which models are offered, what the default is, what changing the model does to existing results, and what happens to models that are retired.Read article
  6. Article 6The AI assistant in Tender NavigatorThe chat assistant on the Tender Navigator website and dashboard: the Pre-Sales Agent for visitors, the Q&A Agent and Full Assistant for subscribers, what each can help with, and how chats are kept.Read article
  7. Article 7Limitations of AI evaluationWhat Tender Navigator's AI features can and cannot do, drawn from its AI Transparency Notice: advisory outputs, language and jurisdiction accuracy, regulatory recency, technical specifications, training bias and hallucination.Read article