2 months ago
Bengaluru, IndiaMid Level
Responsibilities
- Design, automate, and operate end-to-end machine learning training, evaluation, and deployment pipelines using Azure-native tooling.
- Manage containerized model endpoints, model versioning, traffic management, and rollback mechanisms across environments.
- Implement model performance monitoring, data drift detection, observability, and production alerting frameworks.
- Govern LLM provider relationships, API access, versioning, token consumption, and cost optimization across workloads.
- Maintain LLM gateway and prompt-versioning tooling such as LangSmith or Helicone for tracing, evaluation, and production prompt lifecycle management.
- Support reproducible experiment tracking, model cataloging, model registry processes, and governed paths from experimentation to production.
- Translate data science and AI engineering requirements into reliable, scalable production systems.
- Champion CI/CD, infrastructure as code, automated testing, and operational automation for ML workflows.
- Monitor infrastructure spending, identify optimization opportunities, and ensure platform SLAs are met.
Requirements
- 3–4 years of experience in machine learning, Azure, deployment, and pipelines.
- Experience building and operating machine learning pipelines and production model deployment infrastructure.
- Knowledge of model monitoring, data drift detection, observability, versioning, rollback, and model serving.
- Experience with LLM providers, LLM gateways, prompt versioning, tracing, evaluation, or provider governance.
- Ability to collaborate with data scientists, data engineers, and AI architects to operationalize models.
- Understanding of CI/CD, infrastructure as code, automated testing, reliability, and cost optimization for ML platforms.
Tech Stack
Categories
About CUBE
CUBE builds AI-driven regulatory intelligence and compliance automation software that monitors and interprets global regulatory change for highly regulated industries, especially financial services. Its SaaS products—RegPlatform, Oden, Intel, and PoweredBy—help compliance and risk teams act on updates and map obligations to controls; founded in 2011 and headquartered in London, CUBE operates globally and is privately held under Hg ownership.