Senior MLOps Engineer

About the role

We are seeking a Lead ML Data Platform Engineer / Senior MLOps Engineer to build and operate the data and machine learning infrastructure that supports production AI systems.
This role owns the design and implementation of reliable, scalable, and reproducible data pipelines across the machine learning lifecycle. You will work at the intersection of data engineering, MLOps, cloud platforms, and machine learning, partnering with data scientists, ML engineers, platform teams, and business stakeholders.
The ideal candidate has strong Python and data-processing skills, can turn loosely defined requirements into production-ready solutions, and understands how to use model performance and failure analysis to continuously improve dataset coverage and quality.

Key responsibilities

1. Data Pipeline Engineering

Build and maintain scalable pipelines for sourcing, processing, transforming, and delivering structured and unstructured data.

Develop workflows for data ingestion from cloud storage, databases, APIs, file systems, and external platforms.

Design pipelines that support large-scale data processing while maintaining performance, reliability, and cost efficiency.

Implement data lineage, metadata tracking, deduplication, validation, and consistency checks.

Build modular, testable, observable, and reusable pipeline components.

Support both batch and near-real-time data-processing workflows where required.

1. Config-Driven Data Curation and Dataset Management

Design configuration-driven workflows using YAML, CSV, JSON, or similar formats instead of hardcoded logic.

Develop filtering, transformation, sampling, aggregation, and dataset-generation strategies based on business and ML requirements.

Support statistical sampling, stratified sampling, clustering, anomaly detection, and targeted data selection.

Enable reproducible dataset and feature versions with clear lineage from source data to training or evaluation outputs.

Build workflows that allow ML teams to generate and refresh datasets efficiently and consistently.

Identify data gaps and improve dataset coverage based on model performance and business requirements.

1. MLOps and Workflow Orchestration

Build and operate scheduled, orchestrated data and ML workflows on managed cloud and ML platforms.

Use tools such as Apache Airflow, cloud-native orchestration services, or equivalent technologies.

Implement monitoring, alerting, retries, backfills, failure recovery, and operational dashboards.

Troubleshoot failures across storage, databases, APIs, compute environments, models, and external services.

Establish standards for deployment, testing, versioning, observability, and production support.

Automate data validation, model testing, pipeline validation, and release processes.

1. Active Learning and Dataset Improvement

Use model performance, error analysis, and failure patterns to identify data gaps and prioritize dataset updates.

Develop workflows for targeted data collection, re-curation, and dataset refreshes.

Track dataset coverage, completeness, consistency, and distribution across relevant dimensions.

Partner with model owners to ensure that curated datasets address real production failure modes.

Continuously improve dataset quality and coverage rather than relying on static training sets.

1. Technical Leadership and Cross-Functional Collaboration

Define technical standards and best practices for data pipelines, ML platforms, and MLOps.

Mentor engineers and data scientists on Python, pipeline design, cloud infrastructure, and production operations.

Work closely with data scientists, ML engineers, platform engineers, and business stakeholders.

Translate technical and business requirements into clear, scalable data and ML workflows.

Communicate technical constraints, assumptions, risks, dependencies, and delivery timelines effectively.

Contribute to architecture decisions and continuously improve the reliability and maintainability of the platform.

Basic qualifications

Bachelor’s or Master’s degree in Computer Science, Data Engineering, Machine Learning, or a related technical field.

6–10+ years of experience in data engineering, MLOps, ML platform engineering, or a related discipline.

Strong proficiency in Python as the primary programming language.

Practical experience with Pandas, NumPy, data structures, algorithms, and writing clean, maintainable code.

Experience building production-grade data pipelines and ML workflows.

Strong SQL skills and experience with NoSQL databases.

Experience with cloud storage, databases, data warehouses, APIs, and distributed data systems.

Hands-on experience with at least one major cloud platform

AWS: S3, SageMaker, Lambda

GCP: Cloud Storage, BigQuery, Vertex AI

Azure: Blob Storage, Azure ML, Data Factory

Experience with workflow orchestration tools such as Apache Airflow or equivalent platforms.

Experience with Git, CI/CD, automated testing, monitoring, logging, and production support.

Experience designing configuration-driven, reproducible, and metadata-aware pipelines.

Preferred qualifications

Experience with MLflow, Kubernetes, Spark, Databricks, or managed cloud ML platforms.

Experience implementing continuous integration, continuous delivery, and continuous training pipelines.

Experience with statistical sampling, embedding-based clustering, KNN, anomaly detection, or active learning.

Experience designing data observability, validation, and monitoring frameworks.

Experience optimizing pipelines for large-scale data processing and cloud infrastructure.

Familiarity with computer vision and multimedia processing, including

Precision, recall, F1 score, and mAP

Object detection and recognition concepts

OpenCV, Pillow, or FFmpeg

Image and video data processing

Bounding boxes, polygons, masks, and keypoints

Experience supporting retail, physical-store, industrial, or other large-scale operational applications is a plus.

Technical Screening Expectations

Python proficiency is a core requirement for this role. Candidates should be able to solve practical Python and data-processing problems involving functions, data structures, error handling, file processing, and libraries such as Pandas and NumPy.
The screening process may include a basic Python coding assessment to evaluate problem-solving ability, code quality, debugging skills, and practical data manipulation techniques.
How You Work

Autonomous by default: You turn loosely defined requirements into working, production-ready solutions and clearly document assumptions.

Strong problem solver: You can break down ambiguous technical problems and implement practical, maintainable solutions.

Fast ramp-up: You quickly understand unfamiliar codebases, data sources, platforms, and APIs by reading existing patterns and documentation.

Operationally minded: You build systems that are observable, recoverable, testable, and reliable in production.

Configurable and pragmatic: You favor reusable, configuration-driven solutions over one-off or hardcoded workflows.

Fast-paced and adaptable: You deliver incrementally and adjust scope effectively as priorities and requirements evolve.

Collaborative: You work effectively with data scientists, ML engineers, platform teams, and business stakeholders.

Quality focused: You treat reproducibility, lineage, testing, monitoring, maintainability, and operational readiness as core parts of the solution.