BloombergNew York 10022
Full Time

Working closely with ML application teams to design seamless workflows for continuous model training, inference, and monitoring.

||none||Senior Ml Platform Engineer||Bloomberg|| ||||">

Senior Ml Platform Engineer

Full Time Bloomberg, 731 Lexington Ave, New York, NY 10022, United States


The AI Group is the central engineering group responsible for driving Machine Learning (ML) adoption at Bloomberg, with over 200 researchers and engineers working together to provide clients with the best-in-class news, research, market data, and analytics using innovative machine learning technology. We directly impact a wide variety of our flagship products, including news, research, pricing, communications platforms, search and discovery tools. We work on a variety of ML fields, including natural language processing, information retrieval, time series analysis, and recommender systems. 

Within the AI Group, the AI Platform team builds systems to help accelerate the development, deployment, and maintenance of scalable AI services. Our group's mission is to standardize, simplify, and scale the development, deployment and maintenance of all machine-learning-based solutions at Bloomberg built on top of cloud native and open source technologies.

Our team makes extensive use of open source technologies such as, Kubernetes, Kubeflow, KServe, Argo, Buildpacks, and other cloud-native MLOps technologies. From technical governance to upstream collaboration, we are committed to enhancing the impact and sustainability of open source. Some prominent examples are -

  • Cloud Native Buildpacks help us solve the problem around constructing production grade ML environments that are composable and customized to individual use cases, without our application developers ever having to worry about issues like CUDA, Python or library compatibility. We are active contributors to the Buildpacks project and hold TOC and maintainership positions
  • Argo Workflows and Hera power our Model Maintenance Infrastructure which provides continuous training and deployment capabilities for production models in Bloomberg. Our team maintains Hera and has been an advocate for Python and ML focused use cases of Argo Workflows
  • KServe was co-created by Bloomberg to provide production-grade model inference for all of our models from SKLearn based regression models to large language models like BBGPT

In this role, you’ll be expected to interact with global open source project teams and communities. If you have a desire to use, develop, and lead open source software projects, we encourage you to apply. To learn more about our activities in the open source community, head over to our Tech at Bloomberg site.

We’ll trust you to:

While working on the team as an ML Platform Engineer, you will have the opportunity to create a more cohesive, integrated, and managed ML model development life cycle. Typical activities include:

  • Architecting, building, and diagnosing production ML systems
  • Working closely with ML application teams to design seamless workflows for continuous model training, inference, and monitoring
  • Defining and providing strong SLAs around latency, throughput and resource (memory / disk / network / CPU / GPU) usage  
  • Interfacing with both ML experts and platform engineers to understand workflows, pinpoint and resolve inefficiencies, and inform the next set of features for the platforms
  • Collaborating with open-source communities and internal platform teams to build a cohesive MLOps experience
  • Troubleshooting and debugging user issues
  • Providing operational and user-facing documentation

You'll need to have:

  • Experience working with programming languages such as Python or Go.
  • A Degree in Computer Science, Engineering, Mathematics, similar field of study or equivalent work experience 
  • An understanding of Computer Science fundamentals such as data structures and algorithms
  • An honest approach to problem-solving, and ability to collaborate with peers, stakeholders and management