Application Reliability Engineer – Data & AI

Michelin; Maharashtra; On-site; INR 1,000,000 - 2,000,000; Full time

Job summary

Michelin in India is seeking a Data and AI Application Reliability Engineer to own post-deployment health, latency and performance of production AI applications and Medallion architectures.

You will collaborate with feature teams to optimize PySpark jobs, build observability dashboards and implement automated self-healing and production readiness gates. Balance proactive engineering with operational reliability, lead incident post-mortems and ensure high user satisfaction.

Qualifications

• Relevant work experience - 5 yrs

Responsibilities

• Performance Optimization: Analyze and refactor resource-intensive PySpark jobs, queries, and API endpoints to optimize cost, execution speed, and compute efficiency.
• Self-Healing Automation: Develop automated recovery routines, DAG rerun triggers, and data quality checks to minimize manual intervention.
• Engineering Alignment: Partner with Feature Teams and Architects for DoD standards and production readiness gates.
• Production Observability & Monitoring: Design, implement real-time monitoring and alerting frameworks for Medallion pipelines, feature stores, and AI applications.
• Incident Management & Root-Cause Analysis: Lead technical resolution for high-priority production incidents and post-mortems.

Skills

Databricks

Python

PySpark

Azure

Medallion architecture

PowerBI

CI/CD

AI/ML fundamentals

Data science fundamentals

Generative AI & LLM

Tools

Azure Data Factory

Key Vault

Azure DevOps

PowerBI

Job description

Technical Expertise

The Data and AI Application Reliability Engineer you are responsible for owning the post-deployment health, latency, and performance of productionized AI application and Medallion architectures followed for applications.

Acting as the problem-solver for technical issues affecting data pipelines, databases, and deployed AI/ML models, ensuring continuous operation and high user satisfaction.

Partner with core Feature teams to perform root-cause analysis, optimize PySpark jobs, and reduce system debt.

Build automated observability dashboards and self-healing mechanisms for automated failure recovery.

As a part of Job you are required to balance your responsibilities between Proactive Engineering & Automation and Operational Reliability/ Production Health.

Key Responsibilities

• Performance Optimization: Analyze and refactor resource-intensive PySpark jobs, queries, and API endpoints to optimize cost, execution speed, and compute efficiency.
• Self-Healing Automation: Develop automated recovery routines, DAG rerun triggers, and data quality checks to minimize manual intervention.
• Engineering Alignment: Partner closely with Feature Teams and Architects to establish strict Definition of Done (DoD) standards and production readiness gates for new deployments.
• Production Observability & Monitoring: Design, implement, and maintain real-time monitoring and alerting frameworks for Medallion architecture pipelines, feature stores, and AI Applications.
• Incident Management & Root-Cause Analysis: Lead technical resolution for high-priority production incidents, conducting thorough post-mortems to eliminate recurring failure patterns.

Azure Data Engineering Stack

• Proficiency in Databricks, Python, and PySpark.
• Azure (ADLS Gen2, Azure Data Factory, Key Vault, Azure DevOps).
• Hands-on experience with Medallion architecture.

Cloud and DevOps Fundamentals

• Understanding of cloud computing concepts and Services, specifically Microsoft Azure.
• Good handson experience on Python.

Good To Have Technical Abilities

• Understanding of PowerBI reports will be a plus.
• DevOps & CI/CD Fundamentals.
• AI & Machine Learning Fundamentals.
• Data Science Fundamentals.
• Generative AI & LLM Fundamentals.

Behaviour

• Problem Solver: Ability to reverse-engineering complex system behavior and tracking down bugs.
• Automation-First: An instinct to automate repetitive tasks.
• Clear Communicator: Ability to explain technical root causes to non-technical stakeholders clearly.

Relevant work experience - 5 yrs