Log In
Sign Up
About GHGSAT
GHGSat offers greenhouse gas detection, measurement, and monitoring services to industrial and government customers around the world. We use our own satellites and aircraft sensors, combined with third-party data, to help industrial emitters better understand, control, and reduce their emissions.
GHGSat’s capability is unique: The company provides high-resolution, local measurements of atmospheric methane and carbon dioxide concentration from space. This further provides the company means to detect greenhouse gas emitters and to visualize and quantify their emissions.
Responsibilities
- Implement MLOps Architecture on an AWS based cloudaddressing core MLOps components for version control, data management, model training and validation, model deployment, model monitoring and lifecycle management.
- Work collaboratively with data scientists across cross functional teams to understand developed models, contribute to model development and efficiently serve AI models
- Develop robust feature stores, leveraging data from varying data products and building scalable ETL/ELT pipelines to suitably prepare data for model training and serving.
- Support serving ML Models as a Product, managing traffic, inference, drift detection and setting up clear visualizations and dashboards to gauge model performance based on define metrics and KPIs
- Optimize model training and serving for performance and scalability, leveraging distributed computing resources and advanced hardware like GPUs and TPUs on the cloud.
- Implement methods and tools to ensure model interpretability and fairness, addressing bias and ensuring transparency in model predictions.
- Set up proactive monitoring and alerting systems to detect and respond to anomalies in data, model performance, and infrastructure health.
Qualifications & Experience
- 5+ years of relevant experience as an MLOps Engineer, DevOps engineer with MLOps experience, Data scientist, Data Engineering.
- Proven experience in managing and deploying machine learning models in a production environment using AWS or Databricks MLOps offerings.
- Extensive experience with AWS, particularly in using services such as AWS Sagemaker Workflow and Feature store, AWS Lake Formation, Redshift, Sagemaker Feature Store, Eventbridge, AWS Glue and Glue data quality.
- Proficiency with Databricks, and MLOps tools such as MLFlow, AutoML, Databricks FeatureStore, Databricks model serving, Databricks jobs, Delta Live Tables, Unity Catalog and Managing Delta Lakes.
- Experience in a python and SQL based environment including ML/AI modelling concepts, postGIS and track record in engineering Data Science/Computer Vision innovations in the GIS space.
- Strong experience with cloud platforms (AWS, Azure, GCP) and container orchestration (Kubernetes, Docker).
- Strong experience with CI/CD tools and version control systems (e.g., Git, Jenkins, GitLab, Github Actions).
- Knowledge of machine learning frameworks (TensorFlow, PyTorch, Scikit-learn) and their deployment in production environments.
- Experience with Databases and cache: Experienced with DBMS technology, including Postgres, SQLServer, Oracle, MongoDB and MySQL and caching mechanisms such as Redis, Memcached and Varnish
Essential Skills
- ETL/ELT with Apache Airflow, AWS Glue, AWS Data Migration Services, DBT
- Deploying Machine Learning Models with AWS SageMaker and/or Databricks MLFlow
- Experience with Data Warehouse and Data Lakehouse like Delta Lake concepts, AWS Redshift and Databricks
- Proficient with MLOps tools such as ClearML, Weights and Biases, MLFlow, Weights & Biases, Comet, Kubeflow
- Knowledge of distributed computing frameworks (e.g., Apache Spark)
- Experience with container technologies (Docker, LXD, Kubernetes, AKS, ECR)
- Knowledge of monitoring tools and log management systems (e.g., Prometheus, ELK stack) for tracking model performance and infrastructure health.
GHGSat offers a creative and highly motivating work environment We offer competitive salaries, health and social benefits including flex-time and continuing development. We are an open and transparent company, and we are committed to preserving a diverse work environment.
