Senior Machine Learning Engineer

GroundedAI Inc.•Vancouver, BC

8d•Hybrid

About The Position

GroundedAI builds software systems that turn complex, real-world data into reliable insights. Our products operate in challenging, industrial environments where correctness, performance, and maintainability matter just as much as innovation. We work at the intersection of software engineering, geospatial data, and spatial computing. Our focus is on shipping robust systems used by engineers in the field, not prototypes or research demos. We're looking for a Senior Machine Learning Engineer to join our product engineering team and build the models that turn raw underground scan data into structured, trustworthy information. This role is ideal for someone with strong applied machine learning experience, particularly in 3D point cloud segmentation and detection, who also has genuine breadth to contribute to structured extraction from text. This is an applied machine learning engineering role first, with point cloud and spatial data as the core problem space.

Requirements

Strong applied machine learning experience, including real production work, not only research or coursework
Direct experience with 3D point cloud data: segmentation, detection, or reconstruction
Comfort working with noisy, real-world data where the "correct" answer isn't always well defined
Genuine breadth beyond point cloud work, ideally some experience with NLP or structured information extraction
Ability to work independently on problems without an established playbook
Strong software engineering fundamentals: production code quality, testing, and the ability to take a model from prototype to shipped feature

Nice To Haves

Experience with geospatial, geological, or industrial sensor data
Familiarity with point cloud libraries and frameworks (Open3D, PyTorch3D, or similar)
Experience building or fine-tuning LLM-based extraction pipelines
Background in mining, tunnelling, construction, or another physical-world industrial domain
Experience generating or working with synthetic training data for underrepresented data types

Responsibilities

Build models that assist in localization and automated quality checks
Develop models for geological prediction from limited expert input
Build segmentation and detection models for 2d and 3d model spaces
Contribute to structured extraction from technical documents, as your breadth allows
Work with noisy, inconsistent, real-world sensor data rather than clean benchmark datasets
Collaborate with the engineers working on geometry and localization
Contribute to technical discussions on model architecture, training data strategy, and what's realistic to ship near-term versus what needs a dedicated research track