We are Global IT Recruitment specialist that provides support to the clients across UK, and Europe. We have an excellent job opportunity for you.
Job Title: Data & Analytics ML Ops Engineer
Location: North Greenwich, London, SE10 0ES (Hybrid - 2 days per week onsite)
Duration: 12 months contract (235 days)Inside IR35 on FCSA Umbrella
Project Scope:
- Review the existing Data Science model and propose model improvements.
- Review the proposed architecture and propose improvements whilst considering our strategy and costs.
- Document CI/CD for a machine learning models within the controls of TfL Data and Analytics.
- Build a backlog of activity that needs to be completed to deliver the solution.
- As a development team build out the solution.
- Fully test out model deployments and document the process clearly.
- Deploy ML model and ensure it is operational by the expected date.
Requirements:
The ML Ops Engineer will be accountable and responsible for understanding the requirements, ensuring the model is built to production standards, looking at how the model can be deployed, as well as streamlining the processes, automating those processes, and ensuring that we're using the right tools correctly.
Initially the ML Ops Engineer will be responsible for reviewing the D&A Data Science proof of concept. They will need to understand through the D&A Product Owner the requirements and what the output needs to look like. They will then ensure that the model has been developed in a manner that ports to a production environment. They will provide feedback and guidance on any model changes that would be needed to optimise for production deployments.
Once the proof-of-concept phase is over and we move to development the ML Ops Engineer will be accountable for the development and creation of the pipelines needed to deploy the model in to a production environment. Working with the D&A Development team
The ML Ops Engineer will take the model that has been developed by the D&A Data Science team and ensure that it is accessible. The key areas of responsibility are building, deploying, managing and optimising the model in a production environment, to ensure smooth integration and efficient operations.
The ML Ops Engineer is responsible for checking deployment pipelines for ML models and triggering CI/CD pipelines. They will need to monitor these pipelines to ensure all tests pass and that the model outputs are generated and sent to the appropriate location. They will review code changes and pull requests from the D&A Data Science team and take these forwards in a controlled manner.
The ML Ops Engineer should enforce security and data governance best practices to safeguard both the models and the data they process.
The ML Ops Engineer will work to put in place BAU processes that will be adopted by D&A. They will define the process and activity that needs to be undertaken building out a way of working site for the activity. They will identify and implement monitoring tools to ensure response times of the model are within tolerance. Closely work with D&A Data Science Team for model review, run the code refactoring, containerization, versioning to maintain the quality.
The ML Ops Engineer will serve as a conduit between the D&A Data Science team and our TSO support team. Ensuring that any faults in production are rectified within the SLA.
The ML Ops Engineer will also be responsible for engaging with the Senior Product Owner, Data Engineers and Technical Delivery Manager who will be working on different areas of the RUC Secondary ANPR project to identify, understand and define dependencies and integration points between the different areas.
The ML Ops Engineer will be responsible to work independently and in a team environment and be passionate about creating highly scalable, efficient, and easy-to-maintain solutions.
The ML Ops Engineer will be proficient in Azure technologies with specific experience in Python and Scala to ensure that they have the background and experience to take forward the model deployment for RUC Secondary ANPR.
Excellent communication skills are also needed to allow effective translation of sophisticated concepts into easily understood languages across a wide range of stakeholders, including Developers, Build Managers, TSO Support teams, Delivery Managers and BI/Data Architects.
The Authority will be undertaking bespoke development in delivering the end-to-end solution, so the ML Ops Engineer will need to be comfortable interfacing with several bespoke data sources written by other parties, other working areas of the programme and project or with 3rd parties when necessary.
Key Knowledge/Skills:
Ability to balance competing tasks and demands effectively, such as ensuring that all assigned development tasks are prioritised and interdependences are worked through with the rest of the development team.
Effective communication with non-technical stakeholders about complex technical concepts to effectively define and prioritise the features, refine the scope.
Capable at actively listening to, negotiating with and managing conflicts, in order to determine scope and prioritisation for yourself and the team, and to effectively collaborate with stakeholders and other technical roles to identify problems, determine solutions, and effectively manage delivery of an integrated product across multiple development teams and technologies
Capable at continually assessing and improving product processes within their teams, product areas, and on the wider programme to enhance the efficiency and quality of product development, agile practise and product strategy.
Solid understanding of machine learning concepts, techniques and frameworks to enable frameworks to be developed.
Ability to ensure that data scientists can use ML models without having to worry about how they're built or maintained.
Experience:
ML Ops Engineer experience of at least 3 years on projects of a similar size, scale and nature to RUC Secondary ANPR.
Technical experience as an ML Ops Engineer.
Experience of implementing ML models using the Azure stack.
Experience in Python and Scala in relation to ML models.