Research Engineer

• Research and development for our core time series and vision models.

• Designing and implementing training and inference infrastructure (GPU orchestration, experiment management, model serving).

• Bridging the research-to-production gap: turning prototype ML code into deployable, maintainable systems.

• Driving MLOps best practices, including model versioning, monitoring, and quality assurance.

• You have deep experience with time-series or vision models, specifically in an R&D context.

• You've taken messy research code and shipped it as a reliable service—more than once.

• You're comfortable with PyTorch and experiment tracking tools (W&B, MLflow, etc.).

• You've set up and managed training pipelines on cloud GPUs (GCP, AWS, or Azure) and dealt with spot instances, checkpointing, and cost optimization.

• You can design and maintain robust CI/CD and MLOps pipelines.

• You're pragmatic about technical debt: you fix what matters, not what bothers you aesthetically.

• Familiarity with distributed training frameworks.

• You've deployed ML systems in regulated or industrial environments.

• You have strong opinions on MLOps and infrastructure.

• How we can imitate human level logic from protocols, manuals and schematics.

• How we can forecast dynamic systems in zero shot.