Principal Engineer, AI

Principal Ai Engineer

As the Principal AI Engineer, you will act as the technical leader for AI solution design, implementation, and operationalization across Harman's BI, AI, and Data ecosystem. Your primary focus will be defining how AI is applied at scale—ensuring solutions are robust, secure, explainable, testable, and production-ready.

You will lead the development of both prebuilt AI integrations and custom AI/ML solutions, while establishing enterprise standards for MLOps, model governance, and lifecycle management. You will ensure AI solutions are not isolated experiments, but fully integrated, scalable systems built on top of the data platform (Databricks).

What You Will Do

• AI Strategy & Technical Leadership
• AI Engineering Leadership:
• Define best practices for AI solution design, deployment, and lifecycle management.
• Use Case Prioritization:
• Identify high-value AI opportunities and guide their technical execution.
• Standards & Governance:
• Establish standards for model development, validation, deployment, and monitoring.
• AI Solution Architecture & Development
• Define architectural patterns for:
• Batch vs real-time inference
• Feature engineering pipelines
• Model reuse across use cases
• Standardize implementation of common AI solutions:
• Forecasting frameworks
• Classification pipelines
• Anomaly detection frameworks
• NLP/document intelligence pipelines
• Ensure solutions are modular, reusable, and scalable
• Data & Platform Integration
• Data Pipeline Alignment:
• Ensure AI solutions effectively leverage enterprise data pipelines (e.g., Databricks).
• Feature & Data Strategy:
• Guide design of features and data structures required for high-performing models.
• Platform
  
  Collaboration:
• Work closely with Platform Engineers on infrastructure, compute, and scalability.
• MLOps, CI/CD & Lifecycle Management
• Define and enforce MLOps standards using MLflow, including:
• Experiment tracking
• Model versioning and registry
• Promotion workflows (Dev → QA → Prod)
• Co-design CI/CD pipelines with Platform Engineering:
• Automated model testing
• Validation gates before deployment
• Environment consistency across stages
• Establish deployment patterns:
• Batch scoring pipelines
• Scheduled retraining jobs
• Model serving endpoints where needed
• Testing, Validation & Trust
• Define testing frameworks covering:
• Model performance validation
• Data validation and schema enforcement
• Backtesting (especially for forecasting)
• Establish standards for:
• Drift detection (data + model)
• Monitoring and alerting
• Drive adoption of:
• Explainability techniques (SHAP, feature importance)
• Business-level validation (not just statistical metrics)
• Security, Governance & Responsible AI
• Define model governance standards:
• Model approval workflows
• Version control and rollback strategies
• Auditability via MLflow and logging
• Ensure:
• Data access controls and compliance
• Traceability from raw data → features → models → outputs
• Drive responsible AI practices:
• Bias detection and mitigation
• Transparency and explainability where required
• Cross-Functional Leadership & Mentorship
• Technical Mentorship:
• Guide AI Engineers and support broader team development.
• Collaboration:
• Align AI initiatives with Data Engineering, BI, and Platform strategies.
• Stakeholder Engagement:
• Translate complex AI solutions into business value and ensure adoption.
• Innovation & Continuous Improvement
• Technology Evaluation:
• Continuously assess emerging AI tools, frameworks, and capabilities.
• AI Platform Evolution:
• Drive improvements in AI tooling, workflows, and scalability.
• Automation & Efficiency:
• Promote automation and reusable AI components.

What Success Looks Like

• AI solutions are scalable, production-ready, and reusable across use cases
• Models are governed, traceable, and continuously monitored
• MLOps processes (MLflow, CI/CD) are standardized and widely adopted
• AI solutions are deeply integrated into data pipelines and business workflows
• The organization consistently delivers reliable, trusted AI at scale—not experiments

What You Need to Be Successful

• Expert-level Python and deep experience with ML/AI frameworks
• Strong hands-on experience with MLflow (tracking, registry, lifecycle management)
• Deep experience building and deploying production-grade AI/ML systems on Databricks
• Strong experience with…