Role Summary
Visa is hiring a Software Engineer (1-2 years' experience, Python / Go, ML Engineering) for its technology team in Bengaluru, Karnataka (Hybrid - 3 days in office). In this early-career engineering role, you will bridge software engineering and machine learning by developing feature pipelines, training and serving models, and integrating GenAI/agentic workflows. You will build testable production code, optimize ML inference pipelines, construct evaluation suites for non-deterministic LLM systems, and deploy scalable ML services into enterprise production environments.
Key Responsibilities
- ML Pipeline & Software Development: Write, modify, and review clean Python/Go code across data processing, model training, feature stores, and inference layers.
- LLM & Agentic Integration: Build and maintain agentic flows, tool-calling interfaces, multi-step orchestrations, and LLM-powered applications within enterprise workflows.
- Model Deployment & Serving: Package, deploy, and monitor production models using containerization, Kubernetes, and serving runtimes (Triton, TorchServe, or ONNX Runtime).
- Evals & Automated Testing: Develop unit and integration tests along with evaluation/regression suites to measure model behavior, prompt quality, and agent task success rates.
- Observability & MLOps: Maintain automated CI/CD deployment pipelines, monitor live model drift and latency, and resolve production incidents using AI-assisted debugging tools.
Key Qualifications
- Education: Bachelor’s degree in Computer Science, Engineering, Mathematics, Statistics, or a quantitative discipline.
- Experience: 1 to 2 years of professional experience in Software Engineering or ML Engineering.
- Core Languages & Frameworks: Strong programming capabilities in Python or Go; hands-on experience with ML frameworks such as PyTorch, TensorFlow, scikit-learn, or XGBoost.
- GenAI & Agentic Experience: Practical experience with LLM applications, retrieval-augmented generation (RAG), agentic coding environments (e.g., Claude Code, GitHub Copilot), and evaluation frameworks.
- Engineering Best Practices: Familiarity with SSDLC, Git, container technologies, Kubernetes, workflow orchestrators (Airflow/Dagster/Kubeflow), and secure coding standards.
About Visa
Visa is a global leader in payment technology, facilitating secure digital transactions across more than 200 countries while empowering engineers with cutting-edge cloud, data, and AI platform tools.