New opportunity

Lead Software Engineer - ML OPS Engineer

Societe Generale · Bangalore, India

About this role

As a Lead Software Engineer - ML OPS Engineer at Societe Generale in Bangalore, you will be responsible for the deployment and maintenance of the group data science platform infrastructure, enabling data science pipelines to be deployed and scaled efficiently. You will collaborate closely with Data Scientists, Data Engineers, business stakeholders, and the Global Technology Service infrastructure team (GTS) to support end-users and ensure robust model lifecycle management.

Responsibilities

Implement techniques and processes for supporting the development and scaling of data science pipelines.
Industrialize inference, retraining, and monitoring of data science pipelines while ensuring maintainability and compliance.
Provide platform support, understand end-user requirements, and anticipate future development needs.
Collaborate closely with Data Scientists, Data Engineers, and business stakeholders.
Stay updated and demonstrate a keen interest in the MLOps domain.

Required qualifications

Technical skills in Python, FastAPI, SqlAlchemy, NumPy, Pandas, Scikit-learn, and Transformers.
Experience with containerization and orchestration tools including Kubernetes and Docker.
Familiarity with testing frameworks such as Pytest.
Proficiency in CI/CD and DevOps tools such as Jenkins, Ansible, GitHub Actions, Harbor, and Docker.
Strong soft skills including client focus, team spirit, innovation, and a strong sense of responsibility.

Benefits

Opportunities to dedicate several days per year to solidarity actions during working hours, including mentoring, financial education for apprentices, and charity support.
Engagement in the group's ESG strategy and environmental protection practices.

For full details about this role, click “Apply on company site” to view the employer’s original posting.

Skills for this role

PythonFastApiSqlAlchemyNumpyPandasScikit-learnTransformersKubernetesDockerPytestJenkinsAnsibleGitHub ActionHarborAzureDataikuSnowflakePostgreSQLSparkHadoopS3ScalityMAPRML FlowKedroJFROGElastic SearchKibanaGrafanaZabbix

Checking your profile…

Preparation suggestions, not the employer’s actual interview questions. Always check the original posting for current requirements.

Understand the role

What matters most

Data Science Platform Infrastructure Deployment and Scaling
You must focus on building, maintaining, and scaling the group data science platform infrastructure to ensure data science pipelines can be deployed efficiently across different business lines in a hybrid cloud environment using Kubernetes, Docker, and Azure.
Model Industrialization, Retraining, and Monitoring
A core delivery target is industrializing model inference, automated retraining, and pipeline monitoring while ensuring adherence to compliance rules and maintaining high system availability using tools like MLflow, Kedro, and ElasticSearch.
Cross-Functional Collaboration with Data Scientists and GTS
You will need to work closely with Data Scientists, Data Engineers, business stakeholders, and the Global Technology Service (GTS) infrastructure team to align platform features with user needs and support agile ceremonies.
End-User Platform Support and Proactive Evolution
Providing reliable day-to-day platform support to end-users while actively listening to feedback, anticipating future developments, and simplifying complex technical hurdles are essential operational priorities.

Likely challenges to think through

Balancing Standardization with Diverse Business Line Demands
A potential trade-off exists between enforcing rigid, compliant infrastructure standards and accommodating custom data science pipeline requirements across multiple business lines. Prepare by demonstrating how you implement modular infrastructure patterns (using Docker, Kubernetes, and FastAPI) that offer flexibility while maintaining enterprise governance and security.
Managing Complex Hybrid Cloud and On-Premise Integrations
Integrating modern data science tools (such as Dataiku, Snowflake, and MLflow) with legacy or on-premise cloud infrastructure and distributed computing frameworks (Spark, Hadoop) creates operational friction. Prepare by reviewing your experience with hybrid deployment patterns, CI/CD automation via Jenkins/GitHub Actions, and monitoring infrastructure reliability using Grafana and Prometheus or Zabbix.

Practice your answers

Technical practice

Designing Scalable Data Science Platform Infrastructure

What to coverDiscuss how you would design and maintain a data science platform infrastructure to support data science pipelines across various business lines. Address containerization strategies using Docker and Kubernetes, resource isolation, and ensuring high availability and compliance.

Industrializing Model Inference and Retraining Pipelines

What to coverExplain the architectural patterns and tools (such as MLflow, Kedro, or Airflow) you use to industrialize model inference and automated retraining. Address how you ensure reproducibility, traceability, and maintainability in production pipelines.

Building High-Performance APIs with FastAPI and SQLAlchemy

What to coverDescribe how you build robust microservices or platform APIs using FastAPI and SQLAlchemy. Explain how you handle asynchronous operations, input validation, database session management, and efficient querying.

Container Orchestration and Scaling with Kubernetes

What to coverExplain how you manage scaling, resource allocation, and persistent storage for machine learning workloads on Kubernetes. Discuss strategies for handling GPU/CPU resource limits and pod autoscaling during peak retraining or inference loads.

CI/CD Pipeline Implementation using Jenkins and GitHub Actions

What to coverDetail your approach to setting up CI/CD workflows for MLOps using Jenkins, GitHub Actions, and container registries like Harbor. Explain how you integrate automated testing via Pytest and security scanning into the deployment pipeline.

Infrastructure Monitoring and Observability

What to coverDiscuss how you configure monitoring and logging for data science platforms and deployed models using ElasticSearch, Kibana, Grafana, and Zabbix. Explain what key metrics you track to detect data drift, latency spikes, or infrastructure failures.

Big Data and Distributed Computing Integration

What to coverDescribe how you integrate distributed computing frameworks (such as Apache Spark) and big data storage layers (Hadoop, S3, Scality, or MAPR) with machine learning pipelines. Address data transfer bottlenecks and memory management.

Ensuring Compliance and Governance in MLOps Platforms

What to coverExplain how you incorporate compliance, data privacy, and auditing requirements into shared data science infrastructure. Discuss how version control for data, code, and models supports regulatory adherence in enterprise environments.

Behavioural practice

Demonstrating Client Focus with Platform End-Users

What to coverDescribe a situation where you worked closely with non-engineering stakeholders or data scientists who had conflicting technical requests or tight deadlines. Explain how you actively listened, understood their needs, and balanced platform standards with user delivery goals using a situation/action/result approach.

Fostering Team Collaboration and Sharing Expertise

What to coverProvide an example of how you organized collaboration or technical workshops to solve a complex infrastructure or MLOps challenge. Detail how you shared your expertise with colleagues to arrive at the most suitable solution.

Proposing Innovation and Preferring Simplicity

What to coverShare a time when you introduced an innovative solution or streamlined a complex deployment process by prioritizing simplicity over complexity. Describe the trade-offs you evaluated and the impact on the team or platform performance.

Taking Ownership and Meeting Commitments Under Pressure

What to coverDiscuss a high-stakes project or platform outage where you took full ownership to resolve an urgent issue and respect deadlines. Explain how you managed stakeholder communication, diagnosed the root cause, and ensured long-term stability.

Sharpen your knowledge

Topics to prepare

MLOps Pipeline Architecture and Lifecycle Management
Review concepts related to industrializing machine learning model inference, automated retraining, and continuous monitoring. Exercise: Design a conceptual architecture diagram showing how models transition from development using Scikit-learn or Transformers to production scaling on Kubernetes, incorporating compliance checks. Relevance: Directly aligns with the core responsibility of deploying and scaling data science platform infrastructure.
Containerization and Infrastructure Orchestration
Study container management, networking, and scaling principles in cloud and on-premise environments. Exercise: Write a multi-service container configuration using Docker and define Kubernetes deployment manifests with resource limits and health probes. Relevance: Addresses required qualifications in Kubernetes and Docker for platform reliability and scaling.
CI/CD Automation and DevOps Workflows
Review automated testing, build pipelines, and artifact management workflows. Exercise: Draft a pipeline configuration file using GitHub Actions or Jenkins that integrates unit testing via Pytest and artifact storage in Harbor or JFrog. Relevance: Fulfills the requirement for proficiency in CI/CD tools and DevOps engineering.
Distributed Computing and Big Data Environments
Understand how large-scale data processing frameworks handle data pipelines. Exercise: Review data loading and transformation patterns using Apache Spark and distributed storage concepts. Relevance: Directly corresponds to the big data and distributed computing environment specified in the posting (Spark, Hadoop, S3/Scality, MAPR).

Skills to revise

Python Programming with FastAPI and SQLAlchemy
Practice building robust backend services and interacting with databases using Python. Exercise: Build a sample API endpoint using FastAPI and SQLAlchemy that logs model inference requests and validates inputs using Pydantic. Relevance: Directly targets the explicitly required technical stack for platform services.

Your preparation checklist

Prepare an End-to-End MLOps Pipeline Walkthrough
Develop a concrete code or architecture walkthrough showing how you package, deploy, and scale machine learning pipelines using Python, FastAPI, Docker, and Kubernetes. Ensure you can explain how inference and automated model retraining are industrialized and monitored.
Build a CI/CD Automation and Testing Artifact
Create a sample repository demonstrating an automated CI/CD pipeline using Jenkins, GitHub Actions, and Harbor. Include automated unit testing with Pytest and container artifact management to match the required technical workflow.
Design a Scalable Data Science Platform Architecture
Draft a system design diagram illustrating how data science platform infrastructure connects distributed computing tools (like Spark), cloud/on-premise environments, and data sources (such as PostgreSQL, Snowflake, or Hadoop) while ensuring compliance and maintainability.
Compile Evidence of Cross-Functional Collaboration
Prepare professional examples highlighting how you have worked closely with Data Scientists, Data Engineers, and infrastructure teams (such as GTS) to translate user requirements into scalable platform features during agile ceremonies.
Formulate Insightful Questions for the Interviewers
Prepare thoughtful questions regarding the group data science platform's architecture, collaboration workflows with business lines, and how compliance and model lifecycle management are handled across hybrid cloud environments.

Availability last checked: 4 Nov 2026. Status can change after a check; confirm on the employer’s website. Report an issue · Our checking policy

KEEP EXPLORING

Similar AI jobs

Related skills and specializations in India. Matched to this role, not your profile.

Explore more jobs

Finding similar opportunities…