BIMPulse presents ... BIMLogiq (Orange County)

The Evolution of AI in AEC: From AI assistants to AI Agents AI in AEC has evolved significantly over the past few years. We started with chatbots, then moved to AI copilots that could understand Revit and help users accomplish tasks. Around three years ago, BIMLOGIQ took this a step further with BIMLOGIQ Copilot and its own fine-tuned code-generation engine, enabling AI to generate Revit API code and create custom automations. These approaches still have limitations—they generally rely on precise prompts, predefined tools and workflows and require users to define what the AI can do. The next step is AI agents: AI that can reason about a goal, break complex problems into steps, write its own tools, choose and use the right tools, and adapt as it works. This is where Argus comes in. Argus is a fully multi-agentic platform built for AEC, where specialized agents can work together to perform complex Revit workflows. In this session, we'll demonstrate how Argus works and show how you can create your own agents to automate workflows that traditional add-ins and automation approaches simply can't handle. Learning Objectives Identify the limitations of traditional Revit add-ins and rule-based automation compared to adaptive, agent-driven workflows. Recognize how AI assistants interact with Revit and act on the live model. Apply AI-assisted code generation to build simple Revit add-ins that automate repetitive documentation tasks. Compare single-task automation (commands) with multi-agent orchestration, and understand when each delivers the most value. Outline the key components of an AI agent, including instructions, tools, context, and feedback loops; explain how they work together within Revit.