AI Drafting combines selected Knowledge Matching suggestions with the current requirement to produce a draft. A responsible person checks sources, completeness, tone and factual accuracy before approval. Available style and language controls are confirmed in the demo and proposal.
AI Drafting combines selected knowledge blocks with the current requirement to produce a draft suggestion. It is a starting point for review, not a final or automatically submitted answer.
Available style, terminology, language and glossary controls are shown in the demo and confirmed in the proposal. The team checks the resulting tone and terminology against the buyer's requirements.
Available source links are shown with the draft so reviewers can inspect the supporting blocks. Source grounding can reduce unsupported output but cannot remove the risk of an incorrect or incomplete answer.
Bid managers and subject-matter experts review, edit and approve drafts before they move to Submission Check. Logging and approval scope are confirmed for the selected tier before contract.
Yes. Source grounding can reduce unsupported output but does not remove the risk. A responsible person must verify the draft against the requirement and approved sources.
Customer content is not used to train third-party models; the exclusion is recorded contractually. The application database, uploaded documents and backups are held in Switzerland. Model inference uses named, contractually bound interfaces. The processing location and contractual boundaries are disclosed before contract. With the separately available Swiss Inference option, the model environment and inference also run in Switzerland.
The draft is a starting point for human editing. Change logging and any promotion into the knowledge library are confirmed for the selected tier before contract.