Specialist: Ml Ops

**Core Description**:
Responsible to support data science projects and solutions by leveraging service engineering and ML Ops experience to solve a variety of use cases across the Telkom Group and for its customers. Expected to be highly skilled in setting up and supporting a MLOps practice and framework with the ability to design, build and scale MLOps components and services for new and existing use cases across the group in a cloud environment.

**Competencies**:**
FUNCTIONAL KNOWLEDGE**

Relational and non-relational database foundational knowledge; Machine learning operations knowledge base; ML concepts, terms and frameworks; ML engineering knowledge (Python, Scala, Shell scripting, Kubernetes, Fast API, Cloud Run, Docker, PyTorch, Yaml); Cloud ML Ops knowledge (Cloud Source Repositories, Automated Build and Release Management, Container Registry Management and Secure Endpoint Orchestration); ML Ops CI/CD and cloud deployment best practises; Cloud containerisation knowledge, standards and deployment options; Cloud computing and platform management (GCP, Azure, AWS, etc.)

**FUNCTIONAL SKILLS**

Problem solving; Communication (written and verbal); Stakeholder management

**ATTITUDES/ LEADERSHIP COMPETENCIES**

Takes initiative; Hard working; Can do attitude; Innovative thinker; Works well in teams; Works well under pressure; Relishes dynamic/

changing environment; Highly skilled in multitasking; Time management

**Responsibilities**:

- Engage with stakeholders to support the design and delivery of data science projects and solutions.
- Use service engineering and MLOps techniques to solve business problems.
- Responsible for working with a team of MLOps engineers to develop and maintain our cloud-based ML and AI development and production platforms.
- Design, build and maintain MLOps repositories, pipelines and components for cloud-based model processing and serving.
- Support MLOps framework processes for exploratory data analysis and solution development.
- Lead and develop a team of junior MLOps engineers.
- Contribute to our agile way of work and our innovation culture.
- Up to date knowledge of ML platforms and related technologies.
- Translate business requirements into system requirements.
- Consistent documentation of all implemented ML models, systems and processes.
- Support tools, ML models and infrastructure lifecycles via standard service management principles and processes.
- Execute on automation directives by taking repeatable tasks, writing code to replace repeatable tasks, and then adding either a scheduler or some other trigger to enable the job to run automatically while being monitored.
- Enable ML and workload orchestration by configuring and controlling systems that can scale horizontally using specialized tools and techniques.
- Enable ML and component containerization by isolating individual services into containers, allowing them to run anywhere.
- Enabling containers to run and scale horizontally using orchestration tools.
- Execute on ML engineering directives by using engineering techniques to automate and scale the ML model life-cycle and host models in a production environment.
**Required Certification**:
Data and cloud certifications will be advantageous (GCP, Azure, AWS) as well as certifications for other products in our stack (Kubernetes, Istio, FastAPI, Docker, PyTorch, Yaml)

**Qualifications**:
3-year degree/ diploma (NQF level 6) preferably in Computer Science, Mathematics, Statistics, Machine Learning or a related field. A relevant post graduate degree will be an added advantage.

**Experience**:
3-5 years relevant experience, of which at least 2 years must have been in a machine learning operations environment. Experience in ICT/ Telecommunications will be an advantage. Experience with system and process analysis and design.

**Special Requirements**:
Experience with Google Cloud Platform.

Expected to stay abreast of new machine learning frameworks and developments and to put them into practice.

**Comments**: