As an AI Operator, we value not only building the best models for our clients, but also ensuring that these models behave and perform at their highest potential once operationalized. Automatic deployment and continuous monitoring of the deployed models and the underlying infrastructure powering them is a fundamental prerequisite in order for our clients to confidently take full advantage of the predictive and intelligent models we build for them, develop business practices based on the model results, and, therefore, transform themselves into actual AI-driven businesses.
We seek to achieve consistency, repeatability, and speed in the operationalization of all models. This will enable fine tracking of model usage and performance, and provide better options to meet business requirements and risk mitigation objectives.
If you think that POCs are not enough, and that you strive to bring models to production, if you love not only seeing what kind of value data science can bring, but also being a part of building continuous delivery and automation pipelines for the full model lifecycle: we are looking for a passionate ML Ops contributor to move this strategic initiative forward, bring value, and to rise up to the challenge.
As ML Ops, you are in charge of responding to project needs in order to maximize impact on client operations, while building a company-wide ML Ops infrastructure, tooling and methods. You stand at the crossroad of several technical teams: Data Scientists, Software Engineers, Data Engineers and DevOps. You work closely with the Head of ML Ops and the Head of Engineering and Data Science.
- Develop and maintain automated pipelines for model training and consistent deployment.
- Monitor models and their generated outputs continuously;
- Monitor data availability, quality;
- Ensure relevance of model results via monitoring of input data, model output and performance.
- Manage incidents in a timely fashion. Diagnose root cause, and take action to recover service level.
- Facilitate capitalization with feature stores, model catalogs, reporting dashboards;
- Document and leverage past achievements and performance.
- Develop robust solutions to improve overall model deployment and monitoring tooling;
- Anticipate and respond to projects’ needs;
- Leverage relevant open source technologies to accelerate project delivery and contribute to the development of our monitoring and observability infrastructure.
- Watch, screen and evaluate the latest technologies used by the ML Ops community and facilitate their usage at Fieldbox.