DevOps / Cloud Engineering Intern (MLOps & AI/ML Exposure)

About the Internship

VEGAH is looking for a DevOps / Cloud Engineering Intern with working exposure to MLOps and AI/ML to join our technology team. This is a hands-on developer role centered on cloud DevOps on Azure and AWS — building, automating, and operating the infrastructure and delivery pipelines that let data and machine learning workloads run reliably in production.

As a secondary focus, you'll support MLOps workflows — experiment tracking, model packaging, and reproducible deployment pipelines. As a third focus, you'll get practical exposure to AI/ML by working with applied predictive analytics on structured enterprise data (payment delays, supplier delivery risk, customer behavior, service ticket escalation, and operational risk). You'll also develop full-stack awareness — understanding how services, APIs, and lightweight front ends fit together — and you'll be encouraged to use modern AI-assisted vibe coding workflows to move fast while keeping engineering quality high.

This role suits someone who thinks in pipelines, infrastructure, and automation first — and is excited to apply those skills to machine learning systems.

Work Schedule & Working Style (Mandatory)

• Must be available to work during USA CST (Central Standard Time) business hours.
• Readiness to work proactively and autonomously — you will be expected to take ownership of tasks, identify what needs doing, and drive work forward with minimal supervision.
• Be ready for impromptu meetings — the team works in a fast-moving environment where quick syncs, ad-hoc calls, and same-day discussions are common; you should be responsive and flexible.
• Comfortable operating with ambiguity, asking good questions early, and communicating progress clearly and frequently.

Responsibilities

Primary — DevOps & Cloud (Azure / AWS)

• Build and maintain CI/CD pipelines (e.g., GitHub Actions, Azure DevOps) for application and ML workloads.
• Work with cloud services on Azure and/or AWS: compute, storage, networking basics, and managed services.
• Containerize applications and services (Docker) and support deployment and orchestration.
• Automate environment setup, configuration, and repeatable deployments (IaC basics — e.g., Terraform, Bicep, or CloudFormation).
• Support monitoring, logging, and operational reliability of deployed services.
• Use Git/GitHub workflows, branching, and code review practices.

Secondary — MLOps

• Support experiment tracking (e.g., MLflow), model versioning, and reproducible training runs.
• Help package models into reusable components and deploy them through automated pipelines.
• Contribute to model/pipeline packaging and promotion across environments.

Third — AI/ML

• Prepare and analyze structured datasets using Python (pandas, NumPy).
• Build and evaluate classification and regression models (scikit-learn, XGBoost, LightGBM, CatBoost).
• Apply appropriate validation techniques and metrics; identify overfitting, class imbalance, and data leakage.
• Use explainability techniques such as SHAP where appropriate.
• Explore AutoML and newer approaches for tabular data.

Cross-cutting — Full-Stack Awareness & Vibe Coding

• Build simple reusable Python components and APIs (e.g., FastAPI) and understand how they integrate with front ends and internal tools.
• Use AI-assisted development (vibe coding) responsibly — leveraging AI tools to accelerate prototyping, refactoring, testing, and documentation — while reviewing and validating all generated code.
• Document experiments, infrastructure, results, and recommendations.
• Work with VEGAH subject-matter experts to translate business problems into technical use cases.

Required Qualifications

• Working knowledge of Python and basic SQL.
• Hands-on familiarity with cloud platforms (Azure and/or AWS) — compute, storage, and basic networking concepts.
• Understanding of DevOps fundamentals: Git/GitHub, CI/CD concepts, containerization (Docker), and command-line tooling.
• Awareness of the AI/ML lifecycle — data preparation, training, evaluation, packaging, and deployment.
• Familiarity with pandas, NumPy, and scikit-learn; understanding of basic classification/regression concepts.
• Must be available to work USA CST hours, with readiness for impromptu meetings.
• Demonstrated ability to work proactively and autonomously — self-directed, ownership-oriented, and comfortable with ambiguity.
• Strong communication skills and responsiveness in a fast-paced, distributed team.
• Ability to learn new tools and platforms independently.
• Openness to AI-assisted development workflows with sound engineering judgment.

Preferred Qualifications

Experience with one or more of the following is helpful but not required:

• Azure DevOps, GitHub Actions, or other CI/CD tooling
• Terraform, Bicep, or CloudFormation (IaC)
• Kubernetes or container orchestration basics
• MLflow, model registries, or MLOps tooling
• XGBoost, LightGBM, CatBoost, AutoGluon, SHAP
• FastAPI and REST API development
• Light full-stack exposure (simple web front ends, dashboards, integrations)
• Kaggle or personal projects spanning cloud, DevOps, or ML

Students or recent graduates in Computer Science, Software Engineering, Cloud Computing, DevOps, Data Science, Artificial Intelligence, Machine Learning, or related quantitative disciplines are encouraged to apply. Previous SAP experience is not required.

What You Will Work On

Your initial assignment will involve standing up the cloud and CI/CD foundations for benchmarking machine learning approaches against several structured enterprise-data prediction scenarios on Azure/AWS — automating training and evaluation runs, tracking experiments, and packaging the strongest candidates into deployable services with reproducible pipelines.

The objective is to demonstrate which approaches provide the best combination of predictive performance, explainability, development effort, and operational simplicity — and to prove it with automated, cloud-deployed, shippable code.

Strong performers may be considered for future opportunities with VEGAH based on business needs and performance.