HighArc
Research Engineer
About this role
Higharc, a VC-backed startup transforming residential construction with AI, is seeking a Research Engineer to develop foundational spatial AI research combining reinforcement learning, computer vision, LLMs, and 3D geometry. You'll design novel approaches to construction-specific problems, build domain-specialized LLM pipelines, and deliver research-to-production capabilities for the platform.
What you'll do
- Conduct original research in spatial reasoning and architectural structure understanding for residential construction
- Design LLM training pipelines and domain-specific datasets from architectural drawings, building codes, and construction documentation
- Apply computer vision to extract semantic structure from architectural inputs and develop multimodal training approaches
- Develop reinforcement learning methods for generative design and RAG systems grounded in construction knowledge bases
- Build construction-domain benchmarks, ground-truth datasets, and evaluation pipelines for real homebuilding tasks
- Deliver production-ready research outputs including model artifacts, serving pipelines, and API contracts
What they're looking for
- LLM development, fine-tuning, and applied ML systems in production (2+ years)
- Model training frameworks (HuggingFace Transformers, PyTorch)
- RAG architectures and vector databases (Pinecone, Weaviate, pgvector)
- 3D modeling, computational geometry, or computer graphics
- Python proficiency across ML ecosystem
- React/TypeScript for integration work
- Reinforcement learning and computer vision techniques
- Familiarity with BIM tools and data formats (preferred)
Benefits
- Remote-first work with flexible hours
- Competitive salary with significant equity
- Comprehensive medical, dental, and vision coverage
- Flexible PTO and meaningful maternity/paternity leave
- Access to big-company benefits including short and long-term disability
Opens the application — the Jobs AI extension fills it for you. Set up autofill
Opens the official application on the employer’s site. No login required.
HighArc
HighArc is a VC-backed startup building AI-powered software to transform home design and construction workflows for homebuilders. The company is hiring full-stack engineers, solutions engineers, research engineers, and prototype engineers to develop core platform features including showroom experiences, pricing and configuration rule engines, spatial AI capabilities, and research-to-production integrations.
View all jobs at HighArcLikely interview questions
- Walk us through a project where you fine-tuned or trained an LLM on domain-specific data. What was your approach to dataset curation, and how did you measure whether the fine-tuning actually improved performance on your target task?
- Describe your experience building or deploying a RAG system in production. How did you decide between different vector databases, and what strategies did you use to reduce hallucination and ensure grounding in authoritative sources?