ML Ops Engineer

Location : Bengaluru

Experience : 4 - 6 Years

Open Positions : 2

Shift Timings : 1 : 30 PM - 10 : 30 PM IST

Education Qualification : B.Tech

Mandatory Skills :

- Machine Learning

- Python

- Data Warehousing

- AWS

- Data Analytics

- Grafana

Experience Requirements :

- Overall experience of 4 - 6 years, with at least 2 - 3 years of experience in MLOps, ML/Data Engineering, or SRE for data and ML systems.

- Experience collaborating with globally distributed ML Engineering, Model Services, and Tech Ops/SRE teams.

- Flexibility to work with regular overlap with US Pacific Time business hours.

Roles & Responsibilities :

- Monitor model quality in production, including prediction distributions, feature drift, model performance, and data freshness.

- Own dashboards, alerts, and incident triage using Grafana, Datadog, and Prometheus, ensuring documented runbooks for every alert.

- Lead model incident triage across infrastructure, models, and data, and drive Root Cause Analysis (RCA) activities to closure.

- Contribute to Data Quality Monitoring frameworks and maintain training and inference pipelines using Databricks/Spark.

- Support platform migrations, including Lakehouse, Unity Catalog, and runtime upgrades, with validation before and after each migration.

- Improve CI/CD processes and code quality for ML jobs.

- Support model teams through technical documentation, knowledge transfer, and onboarding sessions.

Project Overview :

- The role involves supporting production Machine Learning systems powering personalization, recommendation engines, and search-ranking solutions.

- Models are trained and scored through Databricks/Spark-based batch pipelines. Features are published to feature stores and served through model-serving infrastructure.

- The primary responsibility of this role is to ensure that models continue to perform as expected in production environments, beyond infrastructure availability and system uptime.

- The engineer will focus on :

1. Production model quality monitoring.

2. Model and feature drift detection.

3. Data pipeline health and freshness monitoring.

4. Model-specific dashboards and alerting.

5. Feedback loops and continuous improvement of ML operations.