Applied Research Scientist

Sixone; Burnaby; On-site; CAD 100,000 - 140,000; Full time

Job summary

Sixone in Burnaby, Canada is building end-to-end ML systems that transform spectral and imaging data into actionable recycling decisions. You will work across the stack—from sensor data ingestion and calibration to model deployment in real-world operations.

This hands-on role emphasizes strong ML fundamentals, with development of multimodal models, data pipelines, and edge deployment. You will collaborate with process teams and stay current with spectroscopy, computer vision, and remote sensing

Qualifications

• 2+ years of industry experience in ML engineering or related field.
• Proficient in Python and PyTorch with hands-on model development.
• Experience with high-dimensional data (imaging, spectroscopy, CV, remote sensing).
• Experience with measurement instruments and linking sensor-derived features to physical properties.
• Experience building data pipelines and developing models on cloud infrastructure (AWS).
• Experience in designing ML models (e.g. CNNs, autoencoders) and deploying models on edge devices.

Responsibilities

• Build and maintain data pipelines for cameras, optical spectroscopy, and hyperspectral sensors (ingestion, calibration, normalization).
• Develop algorithms for classification, segmentation, anomaly detection, spectral unmixing and representation learning.
• Design multimodal models linking sensor data to downstream process outcomes.
• Deploy models into production systems.
• Establish data quality and calibration standards for sensors.
• Design and maintain calibration pipelines for imaging sensors.
• Develop preprocessing and feature extraction algorithms.
• Handle variations across sensors, lighting conditions, and acquisition setups.
• Ensure consistency and traceability between raw sensor data and model inputs.
• Establish data quality metrics and validation protocols for sensor outputs.
• Work closely with process teams to translate physical signals into model inputs.
• Own data experimental designs and statistical evaluations.
• Develop data collection requirements and model improvement cycles.
• Stay current with spectroscopy, computer vision, and remote sensing literature.

Skills

Python

PyTorch

High-dimensional data

Cloud AWS

Edge deployment

CNNs

Autoencoders

MLOps

Data pipelines

Tools

Docker

MLflow

CI/CD pipelines

Job description

Sixone’s technology uses machine learning (ML) algorithms to advance our knowledge of consumer plastics. Our goal is to pioneer an ML technology to enable the recycling of complex plastic materials. Sixone has developed technologies for process digitalization and advanced analytics, built to enable efficient plastics recycling. The bridging of our chemical database and sensor readings has been shown to work for blended plastics and plastic-based products. We believe that the economics of recycled materials can be achieved through innovative processing and materials technologies.

Overview of Role

This role focuses on building end-to-end ML systems that transform spectral, hyperspectral, and imaging data into actionable decisions for materials processing. You will work across the full stack: sensor data ingestion and calibration, feature extraction from high-dimensional spectral data, model development and model deployment into real-world recycling operations. This is a hands‑on role that requires strong ML fundamentals.

Responsibilities

• Build and maintain data pipelines for cameras, optical spectroscopy, and hyperspectral sensors (ingestion, calibration, normalization)
• Develop algorithms for classification, segmentation, anomaly detection, spectral unmixing and representation learning
• Design multimodal models linking sensor data to downstream process outcomes
• Deploy models into production systems
• Establish data quality and calibration standards for sensors
• Design and maintain calibration pipelines for imaging sensors
• Develop preprocessing and feature extraction algorithms
• Handle variations across sensors, lighting conditions, and acquisition setups
• Ensure consistency and traceability between raw sensor data and model inputs
• Establish data quality metrics and validation protocols for sensor outputs
• Work closely with process teams to translate physical signals into model inputs
• Own data experimental designs and statistical evaluations.
• Develop data collection requirements and model improvement cycles.
• Stay current with spectroscopy, computer vision, and remote sensing literature.

Candidate Requirements

• 2+ years of relevant industry experience
• Strong proficiency in Python and PyTorch
• Experience in high-dimensional data (e.g. imaging, spectroscopy, computer vision, or remote sensing)
• Experience working with measurement instruments (e.g., spectroscopy, imaging systems, or sensor calibration) and linking sensor‑derived features to physical or chemical properties
• Experience in building data pipelines and developing models on cloud infrastructure (AWS)
• Experience in designing machine learning models (e.g. CNNs, autoencoders) and deploying models on edge devices.
• Comfortable working in a dynamic environment with evolving requirements and continuous product development.
• Effective communication skills with both technical and non‑technical stakeholders.

Preferred Experience

• Experience developing models/ feature extractions from hyperspectral imaging or spectroscopy
• Knowledge in vision transformers / contrastive learning
• Implementing and integrating MLOps (MLflow, Docker, CI/CD pipelines) into workflows.

Sixone offers a stimulating work environment that promotes creativity, curiosity, and innovation. Join the team and contribute to our mission to transform the recycling industry and promote a sustainable future.