Principal Machine Learning Engineer - QuantumBlack - MLOps
Company: McKinsey
Location: Houston
Posted on: May 16, 2022
|
|
Job Description:
You will be part of our global data science and engineering
community and you will lead ML productionization workstreams within
cross-functional Agile project teams alongside Data Scientists,
Machine Learning Engineers, other Engineers, Project Managers,
Translators, and industry experts (i.e., SME).
You will work hand-in-hand with our clients, from data owners,
users, and fellow engineers to C-level executives.
Who you are:
You are a highly collaborative engineer with keen problem-solving
skills and you are passionate about building production-grade
machine learning systems. You know how to engineer beautiful Python
code and enjoy hands-on technical work.
You'll participate in the technical design, development, and
implementation of automated machine learning pipelines in a
production environment using a combination of best-in-class OSS and
vendor technologies. Working within an Agile environment, you'll
serve as a technical lead, providing input into machine learning
architectural design decisions, developing and reviewing model and
application code, and ensuring high availability and performance of
our machine learning systems. Your efforts will be critical to
successfully set up LiveOps protocols.
* Productionize and deploy ML models meeting the high standards and
service-level agreement expectations of production-grade Enterprise
IT in collaboration with multi-disciplinary QuantumBlack teams
* Leverage expertise in modern OSS, Enterprise and public cloud ML
tooling and services to devise the best approach to optimize and
harden production-grade ML models
* Build production ML components such as monitoring pipelines for
data drifts and other quality issues, perform pre-deployment model
performance validation and monitor models in production
* Meet the high bar for software engineering best practices applied
to the data science and analytics space, such as promoting clean
code practices, leveraging static analysis and other code quality
testing and tooling in CI/CD pipelines
* Refactor code for feature engineering, ML model training and
scoring into reusable libraries, APIs and tools to standardize and
accelerate the development-to-deployment lead time
* Play an active role in discussions and workshops for the setup,
deployment, and scaling of use case pilots. Work across analytics
use case squads to enable and ensure adoption and continuous
improvement
* Collaborate with client business leaders from data owners and use
case users to C-level executives to understand their needs and
architect impactful analytics solutions
* Lead and contribute to R&D projects, internal asset
development and the QuantumBlack Engineering community
OUR TECH STACK
While we advocate for using the right tech for the right task, we
often leverage the following technologies: Python, PySpark, the
Python Scientific Stack; AWS SageMaker, GCP Vertex AI, Azure ML for
building scalable ML solutions using public cloud ML stack; Seldon,
Kubeflow, MLFlow, Grafana, Prometheus for machine learning pipeline
management and monitoring; SQL, Airflow, Databricks, our own
open-source data pipelining framework called Kedro, Dask/RAPIDS;
Django, GraphQL and ReactJS for horizontal product development;
container technologies such as Docker and Kubernetes,
CircleCI/Jenkins for CI/CD, cloud solutions such as AWS, GCP,
Azure, Snowflake, as well as Terraform and Cloudformation for
deployment, and many more!
However, we advocate using the right tech for the right task.
Technology evolves and engineering is responsible for staying up to
date with the latest technologies and ensuring we make the relevant
changes where needed.
WHAT YOU'LL BENEFIT FROM:
* Real-World Impact - No project is ever the same. We work with
top-tier clients across multiple sectors, providing unique learning
and development opportunities internationally.
* Fusing Tech & Leadership - We work with the latest technologies
and methodologies and offer first class learning programs at all
levels.
* Multidisciplinary Teamwork - Our teams include data scientists,
engineers, project managers, UX and visual designers who work
collaboratively to enhance performance.
* Innovative Work Culture - Creativity, insight and passion come
from being balanced. We cultivate a modern work environment through
an emphasis on wellness, insightful talks and training
sessions.
* Striving for Diversity - With colleagues from over 40
nationalities, we recognize the benefits of working with people
from all walks of life.
* Continuous development and progression - We offer an extensive
choice of training sessions, ranging from workshops to
international conferences, tailored to your needs as well as a
personal mentorship system. We have multiple career paths and
geographic locations to evolve within the Firm.
* Global community - you will learn from colleagues worldwide by
connecting both internally and externally through our various
hosted meet-ups.
Visit our Careers websiteto watch our video and read about our
interview processes and benefits.
* Bachelor's degree in computer science, engineering or
mathematics
* Demonstrated experience building several high-impact ML solutions
which have achieved meaningful business impact
* Experience setting up at least one contemporary MLOps environment
(e.g., experiment tracking, model governance, packaging,
deployment, feature store)
* Ability to write modern, clean, maintainable, scalable and robust
code in Python and preferably at least one other language (e.g.,
Scala, Java, C++, JavaScript, Bash). Comfort with automated testing
(e.g., unit tests) is a must.
* Applied knowledge of common machine learning algorithms,
techniques (e.g., train/test split, cross validation,
hyperparameter optimization) and evaluation metrics (e.g.,
accuracy, precision, recall, AUC)
* Hands-on expertise with cloud platforms (e.g., AWS, Azure, GCP,
Snowflake), Linux environments, distributed computing frameworks
like pySpark, and using container-based development workflows
automated with DevOps tooling
* Advanced knowledge of SQL and familiarity working with at least
one common RDBMS (mySQL, Postgres, SQL Server, Oracle)
* Ability to scope concrete deliverables with clear milestones and
realistic timeframes
* Experience translating roadmaps to sprint goals while working in
an agile and iterative team set-up to adjust to changes in scope
and objectives
* Passion for and track record of effectively leading and mentoring
more junior colleagues through, such as collaborative code reviews,
pair programming and technical problem solving
* Extensive commercial client-facing or senior stakeholder
management experience
* Willingness to travel in a moderate amount
Keywords: McKinsey, Houston , Principal Machine Learning Engineer - QuantumBlack - MLOps, Engineering , Houston, Texas
Click
here to apply!
|