Senior Machine Learning Engineer

Founded in 2022,
Green Lite is revolutionizing development in America by streamlining the collaboration between developers, builders, and local regulatory authorities. Green Lite’s software powers its Private Plan Review offering, serving many of the nation’s largest public retailers, developers, and production home builders. By leveraging Green Lite’s technology, its customers save months on each project, significantly accelerating their timelines and staying within budget.

Green Lite is founded by experts in technology, development, and within the AEC (Architecture, Engineering, and Construction) industry, and backed by leading venture capital firms. Green Lite is at the forefront of the privatization of construction permitting and plan review, reshaping a multi-hundred billion dollar industry.

Green Lite has raised nearly $40M from the country’s leading venture capital investors, including Craft Ventures, who led Green Lite’s $28.5M Series A. We’re well capitalized to achieve our mission of revolutionizing the plan review and construction permitting process across the country.

We're on a mission to automate one of the most costly and expertise-dependent bottlenecks in the built environment — construction plan review. Today, plan review is slow, expensive, and highly manual, requiring licensed experts to navigate thousands of unique jurisdictional construction codes and complex architectural documents. We believe AI can help.

As a key hire in our AI engineering organization, you’ll operate with a founder mindset, help in defining our long‑term data strategy, and align multiple squads behind it. Our product roadmap includes computer‑vision and large‑language‑model (LLM) capabilities; but those future models will only be as good as the thought that goes into designing them. As an ML engineer you will design, build, and own the way that happens and directly influence company OKRs.

• Work with a rich proprietary dataset encompassing:

  • Access to building codes across every jurisdiction

  • Expert-generated comments on building plans

  • In-house architects and code experts shaping the problem and validating results

• Own the decision making with data engineering on the tech stack we use for MLOps.

• Alongside data engineering, optimize scalable data processing pipelines on our platform and maintain ML infrastructure.

• Model Development:
  Design, develop, deploy and monitor innovative ML solutions that drive efficiency for our customers

• End-to-end Workflow Orchestration:
  Design and maintain complex workflows that automate the path from feature store to inference.

• Design and construct greenfield pipelines that automate pre-training and fine-tuning domain-specific LLM/CV models.

• Define and shape RLHF and retrieval pipelines to inject code-compliance knowledge.

• Evaluations and Experimentation:
  Experiment, evaluate and implement novel research ideas for proprietary LLMs/CVMs

• Have a graduate degree in Computer Science, Data Science or ML related field or equivalent industry experience

• Have 2+ years of experience shipping ML systems (not just research)

• Have experience shipping AI models into production (bonus: experience with AI Agents / LLM orchestration and ML Ops)

• Can efficiently translate open-ended problems into actionable solutions

• Familiarity implementing novel NLP / CVM research ideas and techniques. Prior publications in top conference journals or other evidence of staying current (e.g. open source contributions, conference talks) is a plus.

• Excitement to encode dense regulations & messy CAD & PDF data into structured, learnable signals.

• Prior experience in a startup environment is a plus.

• Excitement for building AI systems that go beyond benchmarks into the real-world messiness of imperfect data

• Thrive in “debate, decide, deliver” cultures—turning ambiguous product goals into concrete, maintainable systems.

90days: LLM/CV model trained on a reproducible pipeline and a clear ML roadmap signed off by product & domain experts.

6months:
Alpha model running in shadow via a fully automated MLOps pipeline, showing>20% quality lift on real…