Principal AI Engineer (Member of Technical Staff)

Role overview A principal-level engineering role focused on the core intelligence layer of a construction-focused AI product. The work centers on multimodal models that ingest blueprint drawings and reason across multiple sheets to surface structured understanding. It is a hands-on, staff-track position with significant ownership over model design, training, and production delivery.

Responsibilities - Design and train multimodal models capable of parsing construction blueprints and architectural drawings. - Build reasoning capabilities that span multiple sheets within a project, connecting entities and annotations across documents. - Take models from research through training, evaluation, and deployment into a production pipeline. - Drive technical decisions on architecture, data strategy, and evaluation methodology for the intelligence engine. - Collaborate closely with product and engineering to ensure model outputs are reliable, fast, and useful for downstream workflows.

Requirements - Deep experience designing, training, and shipping deep learning models in production, ideally multimodal. - Strong background in computer vision and document AI, with familiarity handling structured visual inputs such as drawings or diagrams. - Track record of moving models from prototype to deployed, customer-facing systems. - Ability to reason about data quality, labeling strategies, and evaluation for complex visual reasoning tasks. - Comfort operating at principal or staff level, setting technical direction and mentoring others.

Nice to have - Experience with construction, architecture, engineering, or other technical drawing domains. - Familiarity with retrieval, agentic, or tool-using model patterns that operate over multi-document contexts.