Founding Machine Learning Engineer Full Time

About BuildQ

BuildQ is the AI workplace for project development, finance, and management. We centralize project information, automate key workflows, and provide intelligent tools that make it faster and easier to plan, finance, and build renewable energy projects. Our platform supports teams managing multi-billion solar, wind, and battery portfolios — helping clean energy come online sooner.

We’re a lean team that ships quickly and builds closely with our users. If you want to work on meaningful technology with immediate real-world impact, BuildQ is the place to do it.

The Role‍
We need a Machine Learning Engineer who can build intelligent systems across our entire product stack, owning ML systems that power every stage of a renewable energy deal. This isn’t a research role. You’ll be building production ML systems that customers rely on for multi-million dollar decisions, with clear accuracy targets and customer-facing metrics across the full deal lifecycle.

‍

Who You Are
Must Haves

Shipped ML models to production – handled edge cases, monitored performance, iterated based on user feedback (2-3+ years)
NLP/document understanding background: Experience with LLMs in production: prompt engineering, fine-tuning, RAG systems
Startup 0→1 mentality: high sense of ownership, ability to context-switch between different ML problem types
‍Strong ML fundamentals: Have expertise in Python + ML frameworks, can systematically debug models and design rigorous evaluations

‍
We’re looking for:

You’d rather ship 3 iterations in a month than perfect 1 model in a quarter
You’re comfortable talking directly with customers to understand why a model failed and what “good” actually looks like
You’re energized by ambiguous problems where the “right” approach isn’t obvious – and you know how to make progress anyway
You want to see your work directly impact how users get their job done (not just improve backend metrics)
You’re pragmatic about when to use ML vs. simpler solutions (“let’s use heuristics for V1 and add ML in V2”)
‍