Ptera.ai (previous Vitruv.ai): A Generative Grammar of Classical Architecture

Executive Summary: Vitruv.ai

Slogan: Coding the past for the architecture of the future.

1. Project Vision

Ptera.ai is a Knowledge Engineering platform dedicated to architecture and heritage conservation. The project transforms how professionals interact with historical knowledge, moving from static archives to a generative synthesis engine capable of decoding and applying the "architectural grammar" of different eras.

2. The Problem: Knowledge Fragmentation

The information required for accurate restoration or the integration of new buildings into historical sites is trapped in thousands of disconnected theoretical treatises (Vitruvius, Palladio) and technical execution studies (Orlandos). This lack of a synthesis tool leads to design errors, excessive research costs, and the degradation of urban cultural identity.

3. The Solution: The Ptera.ai Synthesis Engine

Ptera.ai utilizes advanced AI and RAG (Retrieval-Augmented Generation) to create a digital correlation between theory and archaeological practice.

• Module I (The Classical Core): Integrates fundamental treatises with technical archaeological expertise.

• Visual Analysis: The system identifies technical details (e.g., Anathyrosis) directly from on-site photographs or scans, instantly providing historical context and execution rules.

4. Innovation and Uniqueness (The Edge)

Unlike generalist AI models, Vitruv.ai offers:

• Academic Accuracy (Grounding): Every response is mathematically linked to a verified source (specific page, figure, or citation).

• Source of Truth for BIM: We provide the precise architectural data required to fuel modern Building Information Modeling (BIM) and restoration workflows.

5. Validated Technology: The MVP & Controllability

We have successfully validated the concept through a functional Minimum Viable Product (MVP). The system demonstrates the ability to process complex multilingual technical data (Greek, Latin, English). Our core innovation includes a Dual-Mode Precision Layer:

• High-Synthesis Mode (Creative): Used for brainstorming and conceptual synthesis, correlating theory with broad urban contexts.

• Technical-Precision Mode (Zero-Creativity): A strictly grounded mode where the AI acts as a digital philologist. It provides exact measurements—such as the 0.085m anathyrosis margin identified in our tests—eliminating "hallucinations" for critical engineering work.

6. Roadmap and Objectives

Following the successful MVP validation, the project is moving toward scaling the proprietary platform architecture, expanding the indexed database, and launching the specialized visual analysis module for restoration professionals.