Production Machine Learning Engineer, Applied Research

VenorTalent; Halifax; On-site; CAD 80,000 - 110,000; Full time

Job summary

A leading AI solutions provider based in Canada is seeking an experienced Production Machine Learning Engineer to design, deploy, and maintain production-ready models in a sustainable environment. The role entails working with cross-functional teams to develop robust data workflows and improve production systems for aquaculture. Ideal candidates should have 4+ years in applied ML, proficient in Python and PyTorch, and capable of managing interpretability methods. This position promotes a culture of belonging and inclusivity.

Qualifications

• 4+ years in applied ML/data science with shipped production systems.
• Comfortable with image/video models and real-time inference pipelines.
• Ability to analyze layers, features, and learned representations.

Responsibilities

• Design and train image/video models.
• Build evaluation frameworks beyond standard metrics.
• Implement drift detection and distribution-shift tests.
• Collaborate with cross-functional teams for deployments.

Skills

Python

PyTorch

scikit-learn

OpenCV

Data analysis

Interpretability methods

Deployment to edge environments

Monitoring for data drift

Clear communication

Independent execution

Job description

Production Machine Learning Engineer, Applied Research

Venor is thrilled to partner with ReelData AI to identify a Production Machine Learning Engineer (Applied Research). ReelData builds edge-based AI systems for fish farms that help produce healthier fish more efficiently, reduce waste, and strengthen global food security. With camera hardware deployed in water and models running on-device, the work you do directly impacts sustainability outcomes in real-world environments.

This role blends hands-on engineering with applied research to design, deploy, and maintain production-ready models—delivering technology that meaningfully benefits people and the planet.

What You’ll Do

• Design and train image/video models (e.g., transformers, CNNs, autoencoders).
• Build evaluation frameworks beyond standard metrics (residual analysis, calibration, reliability curves).
• Implement drift detection (data/feature/prediction), distribution-shift tests, and canary checks.
• Apply production interpretability/debugging (dataset & label audits, slice discovery, attribution, counterfactuals).
• Build and maintain robust data workflows for ingesting, labeling, and cleaning image/video at scale.
• Run active-learning loops and targeted data curation for narrow tasks.
• Develop autonomous pipelines using Celery, Apache Airflow, and related orchestration tools.
• Establish monitoring and alerting for data quality, drift, and performance.
• Collaborate with hardware, software, and field teams to deploy and maintain systems; occasional travel to customer sites.
• Redesign critical parts of the stack where it drives significant improvement.

What You Bring

• 4+ years in applied ML/data science with shipped production systems.
• Strong Python & PyTorch; comfortable with scikit-learn and OpenCV.
• Experience with image/video models and real-time inference pipelines.
• Hands-on use of interpretability/validation methods (residuals, calibration, attribution, counterfactuals, dataset-level audits).
• Ability to analyze layers, features, and learned representations.
• Proven deployments to edge/resource-constrained environments.
• Familiarity with monitoring for data/drift/performance/system reliability.
• Clear communication with technical and non-technical stakeholders.
• Independent execution with strong ownership.

Nice to Have

• GStreamer and NVIDIA DeepStream.
• Stereovision background.
• Experience in aquaculture, agriculture, or other sensor-based real-world systems.

Why ReelData AI

• Series A company with production systems and an international customer base.
• Opportunity to set the standard for production ML in aquaculture.
• Flexibility to redesign components where it creates the most impact.
• Apply cutting-edge interpretability techniques to high-impact, real-world deployments.

Our Commitment to BelongingAt Venor and Verge Technologies, we embrace a culture of belonging in the workplace. No matter who you are, where you’re from, how you think, what you believe in, or who you love, we welcome your application. We all come from different backgrounds and different walks of life, bringing unique perspectives and experiences. We encourage applications from 2SLGBTQ+, Black, Indigenous, and People of Colour (BIPOC), women, newcomers to Canada, and people with disabilities.

How to ApplyTo apply, please forward your resume to Vanessa Rodriguez Garcia at [e-mail].