Deterministic Blackboard Pipelines with Specialized LLM Knowledge Sources: A Generalizable Architecture for Intelligent Multi-Stage Reasoning — IAIT2026 Presentation Slides

Key Points

• The aim is to present a generalizable architecture for intelligent multi-stage reasoning via deterministic blackboard pipelines using specialized large language models.
• Developed an architecture integrating specialized knowledge sources into deterministic pipelines.
• Utilized multi-stage reasoning to enhance decision-making capabilities.
• Presented findings during the IAIT2026 session.
• Demonstrated improved reasoning effectiveness in complex tasks.
• Validated the architecture's adaptability across different applications.