Washington, D.C. / Philadelphia hub · Hybrid (3 days/week on-site) · Full-timeCompensation: $190,000–$200,000 + competitive equity
About the Company
Our client is building the AI-powered workflow layer for construction — one of the world's largest industries, and one that skipped the software era entirely. Subcontractors, its core customer, have seen no meaningful software innovation in 15+ years, and agents plus multimodal AI are the first real wedge in. Off-the-shelf foundation models struggle with construction diagrams because those aren't in standard training data, which makes training in-house models a core technical bet. It recently closed a seed round and is at ~$2M ARR.
Founded 2023 · 11–50 people (Seed) · Industry: Construction Tech / AI
The Role
The client is hiring a Founding AI/ML Engineer with deep computer-vision expertise to own the model-development side of the platform and lead the in-house model work that off-the-shelf foundation models can't handle.
Tech stack: Python, PyTorch, CUDA; computer vision (2D segmentation, object detection, classifiers; SAM / SAM3-tier foundation models); agentic orchestration & retrieval (LLMs, RAG, fine-tuning small language models, local vs foundation-model trade-offs).
Requirements
- Computer vision, with 2D segmentation depth
- Python + PyTorch + CUDA
- Trains in-house models, not just fine-tunes
- Construction diagrams / floor plans a bonus
- East Coast (D.C. / Philadelphia hub), 3 days hybrid
Nice to Haves
- 2D-segmentation expertise with published work in floor-plan localization, raster-to-sequence, structural priors, or comparable architectural CV
- Personal or professional exposure to construction diagrams, floor plans, or technical drawings — genuinely rare and high-signal
- PyTorch + CUDA fluency with an architectural-model-improvement track record (not just fine-tuning APIs)
- Research-trained but now hands-on
- Agentic orchestration / retrieval experience as a secondary skill
Why Join
- Own the model layer end to end: the founding AI/ML hire, owning the entire model-development side of the platform and setting technical direction
- A genuinely hard, defensible problem: construction diagrams aren't in any foundation model's training data, so you train in-house models — real CV depth, not API wrapping
- Early-stage upside: a meaningful early-employee grant, a fresh seed round, ~$2M ARR with a credible 10x plan over the next 12 months
Details
- Location — Washington, D.C. / Philadelphia hub
- Work policy — Hybrid, 3 days/week on-site
- Compensation — $190,000–$200,000 + competitive equity
- Visa sponsorship — Available (H-1B)
- Employment type — Full-time