Your schematics, your terms
Architecture and engineering firms are rightly cautious about sending sensitive schematics to cloud services. AedifAI supports deployment models to match, from cloud AI for speed to a fully air-gapped install for schematics that can't leave your network at all.
Hosted
For teams that want shared projects, review history, and multi-user collaboration, with the same analysis pipeline running server-side.
Air-gapped
Run your own vision-capable inference endpoint on your network (vLLM, TGI, or similar). No schematic data reaches a third-party AI provider at all. Built for firms and regulated buyers who can't send schematics off-site under any deployment.
Hybrid
No shared backend: project data, findings, and history stay local via OPFS and IndexedDB instead of a hosted database. Analysis itself still calls out to an AI endpoint, cloud by default, or on-network if paired with air-gapped mode.
Have a security questionnaire?
Talk to us directly. We'll walk your security and compliance team through the architecture and deployment options available for your requirements.
Talk to Us