Responsibilities
- Develop the overarching enterprise strategy for AI implementation and long-term evolution.
- Create and maintain a reference architecture that guides AI system development across the organization.
- Define technical and governance standards for building, deploying, monitoring, and managing AI systems.
- Design service layers enabling both immediate real-time decisions and delayed batch processing.
- Ensure AI infrastructure supports deployment across multiple regions, including compliance with data residency laws in the US and EU.
- Serve as principal architect for AI projects spanning various business units.
- Build systems that deliver predictive insights and real-time intelligence through advanced analytics.
- Design AI solutions for predictive maintenance and anomaly detection using IoT data streams.
- Develop event-triggered AI workflows that integrate directly into operational processes.
- Implement data protection measures including encryption, masking, and consent-based processing for sensitive information.
- Integrate model explainability and continuous monitoring into all production AI systems.
- Lead the transition toward agentic AI systems capable of autonomous actions.
- Design AI agents that can reason, use tools, and interact across platforms under supervision.
- Establish operational boundaries, identity controls, escalation paths, and full traceability for AI agents.
- Ensure AI agents function reliably and securely across platforms such as Salesforce, Brightree, Oracle, Five9, ServiceNow, and internal systems.
- Develop testing protocols, adversarial evaluations, and behavioral assessment standards for AI agents.
- Advance the organization's approach to using multiple foundation models.
- Define policies for selecting, integrating, and governing foundation models from providers like OpenAI and Anthropic.
- Design intelligent routing, fallback mechanisms, and cost-optimized inference strategies across models.
- Implement Retrieval-Augmented Generation (RAG) systems with controlled access and knowledge governance.
- Introduce standardized evaluation frameworks to assess model accuracy, grounding, and hallucination rates.
- Avoid dependency on single vendors by designing modular, interoperable AI architectures.
- Establish robust MLOps and LLMOps practices for AI lifecycle management.
- Define governance policies for moving AI models from experimentation to production.
- Implement model registries, automated pipelines, monitoring systems, and rollback capabilities for AI deployments.
