I develop advanced intent-classification systems that act as a central nervous system for AI deployments. AIRS (AI Intelligent Routing System) seamlessly triages user queries from Open WebUI and orchestrates automated responses across multiple business logic paths.

The Workflow Logic:

• Intent Classification: Real-time analysis of user queries to determine the optimal processing path.
• Dynamic Persona Injection: For personalization requests, the system generates a custom System Prompt returned to Open WebUI. This prompt is saved globally, ensuring all future interactions are tailored to that user's specific profile.
• Automated Support Escalation: Critical system feedback is automatically routed to human support teams with full conversation context.
• RAG-Powered Knowledge Retrieval: Business queries trigger high-speed API lookups within the Open WebUI Knowledge Base for grounded, factual answers.
• Ambiguity Resolution: Proactively asks for clarification when user intent is unclear, ensuring high-quality outputs.

Technical Highlights:

• Zero-Shot Intent Classification: Uses advanced LLM prompting to categorize user requests with high precision without needing large training datasets.
• Open WebUI API Deep Integration: Leverages the Open WebUI REST API to dynamically modify User Settings and System Prompts, creating a "Self-Configuring AI" experience.
• Context-Aware Global Prompts: The system doesn't just answer; it reprograms the user's environment by injecting persistent context that follows them across all future chat sessions.
• Conditional Routing Logic (n8n): Features complex switch-case nodes in n8n to handle 5+ distinct logic branches, including external notification triggers and internal database lookups.
• Support Ticket Integration: Bridges the gap between AI and Human Ops by transforming user "System Feedback" into actionable support notifications.