Software Engineering Intern- AI/ML Backend (Spring 2026)

Digs•Vancouver, WA

18h•$20 - $25•Hybrid

About The Position

Digs is building the future of collaboration between builders and homeowners. Our platform uses artificial intelligence to organize, surface, and share the information that matters most during a home build—making the entire experience more transparent, connected, and intuitive. We’re a team of experienced entrepreneurs, operators, and builders who care deeply about craft. We move fast, think long-term, and hold a high bar for what we build. If you’re curious, motivated, and excited about creating meaningful technology from the ground up, you’ll fit right in. At Digs, you won't be working on throwaway sandbox projects. We expect you to hit the ground running and ship real production code from day one. You will act as a core member of our R&D team, working alongside our Head of R&D and the broader engineering team to build the services and infrastructure that turn ML model outputs into user-facing value — focusing on backend integration, evaluation, and production readiness rather than model research. Want to know what drives us? Explore our Digs Principles.

Requirements

You recently completed a degree in Computer Science, Machine Learning, Data Science, or a related field (or have equivalent practical experience) and are looking for a runway to transition into a full-time role.
You have a strong foundation in Python and are comfortable writing production-quality code, not just notebook prototypes.
You have coursework or project experience in machine learning and understand the fundamentals of training, evaluation, and iteration.
You are comfortable working with data — cleaning it, analyzing it, visualizing it, and drawing conclusions from it.
You are a highly collaborative, self-motivated communicator who thrives in ambiguity, takes ownership of your work without waiting for instructions, and isn't afraid to say "I don't know yet" while moving fast.
You are comfortable participating in virtual meetings and in-person collaboration at our Vancouver, WA HQ

Nice To Haves

Experience building APIs with FastAPI, Flask, or similar Python web frameworks and deploying models behind them.
Familiarity with HuggingFace Transformers and the model hub ecosystem.
Experience training models from scratch or fine-tuning using supervised fine-tuning or parameter-efficient fine-tuning techniques.
Contributions to open-source projects or published research.
Startup experience: proven ability to move fast on a small engineering team.
Experience or interest in the home building, home remodeling, construction, or real estate industries.

Responsibilities

Build and maintain FastAPI services in Python that serve ML models and integrate into our broader microservices architecture.
Build evaluation frameworks and benchmarks to systematically measure quality across our ML services.
Work with geometric and spatial data using libraries like Shapely, NumPy, and OpenCV to process, validate, and post-process model outputs into production-ready formats.
Improve and extend our vision and language pipelines for construction document understanding, including tasks such as instance segmentation, classification, and geometric shape analysis.
Integrate ML outputs into production workflows orchestrated by our Restate durable execution engine.
Adopt and help shape our agentic coding practices — our team embraces tools like Cursor, Copilot, and Claude Code to accelerate development.
Work with vector embeddings and semantic search infrastructure using pgvector.
Contribute to our TypeScript backend and React frontend as features require end-to-end implementation.
Run data analysis and experimentation to identify model failure modes and drive improvements with measurable results.
Contribute to our human-in-the-loop annotation workflows, helping design pre-annotation pipelines and process crowdsourced labeling results.