Abstract
A system is described herein for adapting artificial intelligence (AI) assistant landing-page prompts to individual user maturity. A persisted behavioral profile establishes an explorer, practitioner, or power-user stage that controls visible suggestion categories and dynamic-generation eligibility. Before the user submits a query, a coordinator assembles a context bundle from user query history, organization-level intent signals, role-based access control (RBAC) scope, and live network alerts or configuration changes. A large language model (LLM) generates up to five grounded suggestions, which are schema-validated and deterministically filtered for authorization before delivery. Refresh heuristics, context-fingerprint deduplication, per-user concurrency control, and honest empty results bound generation cost and reduce alert fatigue while enabling proactive, network-aware suggestions at assistant load.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Villeneuve, Mitchell; Bykampadi, Pavan; and Nail, Katie Louise, "MATURITY-ADAPTIVE DYNAMIC AI PROMPT GENERATION USING NETWORK-GROUNDED CONTEXT BUNDLES", Technical Disclosure Commons, (September 07, 2026)
https://www.tdcommons.org/dpubs_series/11611