ML Engineer
Build ML systems that learn from layouts and enhance spatial intelligence, recommendations, and 2D→3D automation.
About The Role
You will develop the machine learning layer that enables InHouse to automate and enhance interior layouts, recommendations, and spatial reasoning. This includes building models that learn from designer-created rooms, extracting spatial and stylistic patterns, and evolving the ML components that support our 2D→3D automation. You’ll design training pipelines, evaluate commercial and open-source models, and integrate ML outputs into our backend services. This is a hands-on role spanning applied ML and practical data engineering, collaborating closely with engineering and design.
What You'll Do
Develop ML models that learn from designer-created layouts, including placement patterns, object co-occurrence, scale, and spatial relationships
Build structured scene representations and data pipelines to support ML training and evaluation
Own and evolve ML components supporting 2D→3D automation (layout interpretation, geometry inference, spatial segmentation)
Fine-tune or adapt open-source and commercial models to InHouse-specific data and workflows
Design and implement pipelines for dataset creation, labeling, feature extraction, and model retraining
Prototype heuristics or rule-based systems as stepping stones to full ML automation
Integrate ML outputs into backend services powering layout generation, recommendations, and design guidance
Create evaluation frameworks and metrics to measure model performance, reliability, and improvement over time
Document data structures, model behavior, assumptions, and integration logic for future engineering growth
What You Need
3–6 years of applied ML engineering experience delivering in a production environment
Strong Python engineering ability and experience with PyTorch or similar deep learning frameworks
Experience working with structured or semi-structured data and turning it into features or training signals
Ability to design lightweight ETL/data pipelines supporting ML workflows, including familiarity with embeddings, compatibility scoring, or ranking models
Comfort breaking ambiguous product questions into tractable ML problems
Ability to evaluate OSS, hosted ML models, and research approaches to identify pragmatic solutions
Nice to Have
Experience with scene graphs, spatial reasoning models, layout understanding, and 3D concepts (coordinate systems, object transforms, bounding boxes) with a background in computer vision for geometry inference
Benefits
Hourly contract rate
Contract to full-time opportunities
Quarterly paid travel and accommodation in New York City (from US origin)
Standard contractor benefits where applicable
Other
InHouse is an equal opportunity employer. We celebrate diversity and do not discriminate on any protected basis.
About InHouse
InHouse transforms how renters furnish their homes. We build a photorealistic digital twin of your space from a floor plan and photos. Inside, you can visualize complete rooms, swap products instantly, and shop exclusive pricing from hundreds of premium brands—all guaranteed to fit.
Our platform combines a live multi-brand catalog, spatial-placement engine, and ML-driven tooling to deliver professional-quality interiors ~98% faster and ~95% less costly than traditional design. We're making professional furnishing accessible on any budget, timeline, or skill level.
InHouse is backed by a diverse group of venture, angel, and strategic investors. The founding team brings 40 years of experience in e-commerce, AI, and design across 10+ venture-backed startups.
Why Join InHouse?
InHouse is redefining how people design and shop for their homes by merging photorealistic visualization, ML-assisted workflows, and seamless commerce into one cohesive platform. This is a high-impact opportunity to design the intelligence layer that shapes how spaces are laid out, how products are recommended, and how the system learns over time. You’ll have autonomy to define ML direction, experiment with models and data, and bring research-backed ideas into a real product environment. Your work will directly influence user experience, reduce reliance on manual design workflows, and unlock new capabilities across the platform, from product to ops.
Submit your application.
We’ll be in touch once we've reviewed your application.
