Director, Applied Science (Responsible AI​/Multimodal

Position: Director, Applied Science (Responsible AI/Multimodal)
** Job Description*
* Oracle Cloud Infrastructure is seeking a  
** Director of Applied Science
** to lead the Responsible AI Applied Science team. This leader will define and execute the scientific and technical strategy for building the next generation of Responsible AI technologies that enable customers to safely and reliably deploy  
** multimodal generative AI models and agentic systems
** at enterprise scale.

The Responsible AI Applied Science team advances research and develops cutting-edge technologies for  
** guard railing multimodal generative AI models and AI agents** . The team's work spans models and applications that reason over and generate across modalities-including  
** text, images, audio, video, and combinations of these modalities** . Research areas include content safety and moderation, prompt injection and adversarial attacks, privacy and PII protection, hallucination and grounding, reliability, formal verification, and emerging approaches for governing and constraining agent behavior.

As generative AI becomes increasingly multimodal, safety challenges extend beyond individual prompts and responses to the interactions between modalities, models, tools, and agents. The team develops new techniques for understanding and mitigating these risks while building guardrails that can operate effectively across diverse models, modalities, and enterprise use cases.

This is a senior scientific leadership role at the intersection of  
** research, product, and engineering** . The Director will lead a team of applied scientists, establish the team's research and product roadmap, and translate advances in multimodal Responsible AI into differentiated capabilities for OCI customers. The successful candidate will balance breakthrough research with pragmatic execution, ensuring the team's work results in production technologies, high-impact publications, patents, and meaningful advances in the state of the art.

** Responsibilities*
* ** What You'll Do*
* ** Set the Responsible AI Science Strategy*
* + Define the Applied Science roadmap for  
** multimodal Responsible AI
** in OCI, identifying the highest-impact scientific and customer problems for the team to pursue.

+ Establish priorities across near-term production deliverables and longer-term breakthrough research.

+ Develop a research agenda for guard railing generative AI across  
** text, vision, audio, video, multimodal models, and agentic systems** .

+ Anticipate emerging risks and technical challenges as generative AI evolves from single-modality models toward increasingly multimodal, autonomous, tool-using, and agentic systems.

+ Partner closely with Applied Science Architects, the Chief Scientist, engineering leaders, and product leadership to identify and invest in breakthrough science opportunities that can create differentiated capabilities for OCI.

** Lead a High-Performing Applied Science Team*
* + Lead, mentor, and grow a team of applied scientists working across Responsible AI, multimodal AI safety, and AI reliability.

+ Establish a culture of scientific rigor, technical ambition, experimentation, collaboration, and customer impact.

+ Provide technical and scientific direction while enabling senior scientists to independently lead significant research areas.

+ Attract and develop exceptional scientific talent and build the capabilities required as the Responsible AI and multimodal AI landscapes evolve.

** Advance the State of the Art in Multimodal Responsible AI*
* Drive research and innovation across areas including:

+ Multimodal content safety, moderation, and policy enforcement across text, image, audio, and video

+ Multimodal prompt injection, jailbreaks, and adversarial robustness

+ Privacy, PII, and sensitive-information detection and protection across modalities

+ Hallucination, grounding, and factuality detection and mitigation

+ Safety and reliability of multimodal foundation models

+ Cross-modal safety, including risks that emerge from interactions between different input and output modalities

+ Formal methods and verification for AI systems

+ Evaluation, benchmarking, and measurement of Responsible AI systems across models and modalities

+ Guardrails for AI agents, including tool use, planning, actions, multimodal inputs and outputs, and multi-agent behavior

+ Novel approaches to runtime monitoring, intervention, and enforcement for multimodal generative AI systems

Identify opportunities where advances in these areas can become  
** production capabilities, publications, patents, or strategic intellectual property for Oracle** .

** Translate Research into Customer Value*
* + Work closely with engineering, product management, and other OCI stakeholders to product ionize Responsible AI research into  
** scalable multimodal guardrails and services** .

+ Develop a deep understanding of enterprise customer requirements and translate them into meaningful scientific and technical problems.

+ Build guardrail technologies that can operate across different generative AI models…