Principal Jobs in San Francisco, CA

About Okta Secures AI Team

The Okta for AI Agents Team is building the future of digital identity management. The way companies are using agents and allowing them access is undergoing a fundamental shift, enabling teams to move fast while staying secure.

Our North Star is clear: To get AI right, you have to get identity right. We are not just managing identities; we are building the industry's first Identity Security Fabric for the agentic era. As organizations rapidly deploy AI, these non-human entities are acting with access to critical tools, often bypassing traditional perimeters and creating a 'Shadow AI' crisis. Our mission is to move identity from a reactive gatekeeper to a unified control plane, transforming AI risk into business ROI.

We are tackling this by implementing a comprehensive four-pillar maturity model: Discover, Onboard, Protect, and Govern. We are driving innovation by defining the standards for Agentic Identity—including pioneer work with securing AI Agents and support for protocols like MCP. You will be helping us build the infrastructure that allows enterprises to scale AI safely, ensuring every access decision—whether made by a human or an automated agent—is authenticated, authorized, and audited at scale.

We are a diverse team of engineers, product managers, and designers who are bringing Okta's expertise in identity to define the future of Agent Identity management, focusing on enablement while staying secure.

About the role

In this role, you will join the Agent Access Policies sub-team under Okta Secures AI as a Principal Machine Learning (ML) Engineer to advance Okta's authorization capabilities. By replacing static, rule-based systems with dynamic AI security mechanisms, you will deliver real-time threat inspection, agent intent evaluation, and behavioral analysis—ensuring autonomous AI agents operate safely within specified bounds.

Your core mission will be architecting and driving features for our unified control plane to establish the premier Identity Security Fabric for the agentic era. In doing so, you will help execute our guiding principle: "To get AI right, you have to get identity right."

What you'll be doing

• Implement intent-based enforcement to verify agent runtime requests match their intended purpose, requiring key capabilities.
• Apply LLM reasoning and prompt parsing to interpret prompts, tool payloads, and intent in real time.
• Integrate low-latency inference or semantic evaluation engines directly into the API gateway request path.
• Use embeddings, vector search, or zero-shot classification to score alignment between agent intent and executed actions.
• Design confidence-scored decision engines that feed semantic verification results into policy frameworks (e.g., Cedar).
• Establish evaluation benchmarks, prompt injection defenses, and guardrails to prevent bypasses or false positives.
• Architect scalable ML and Generative AI systems integrating retrieval, inference, and evaluation pipelines.
• Optimize prompting, context retrieval, and RAG workflows for accuracy, safety, and efficiency in Claude-based systems.
• Build automated evaluation pipelines to measure model quality, correctness, groundedness, and safety in production.
• Implement schema validation, structured output enforcement, and guardrails to ensure reliable, compliant AI outputs.
• Mentor and coach engineers to support team and community growth.

What you'll bring to the role

• 10+ years of software development experience, with strong programming expertise in Python (and familiarity with Go or Typescript a plus).
• Hands-on experience with applied machine learning, from feature engineering to training and fine-tuning models.
• Hands-on experience with modern Generative AI platforms (AWS Bedrock, OpenAI, Anthropic, etc.).
• Deep understanding of retrieval-augmented generation (RAG), embeddings, and knowledge-base workflows.
• Hands-on experience with LiteLLM, LangGraph, LangChain, LlamaIndex, MCP, or other related AI agent frameworks.
• Familiarity with ML frameworks (FastAPI, PyTorch, TensorFlow, Spark ML) and workflow orchestration tools (Airflow, etc.).
• Experience defining evaluation metrics, pipelines, and feedback loops for ML/GenAI systems.
• Proven ability to collaborate with product and engineering teams to drive greenfield initiatives forward, navigate unknowns, and iterate quickly and frequently.
• Experience building tools or infrastructure for AI/ML applications, with a deep understanding of the developer lifecycle in an AI-native world.

Extra credit

• Experience integrating AI-driven systems with identity, authentication, or security products.
• Exposure to ethical AI, model risk, or compliance frameworks.
• Familiarity with evaluation datasets, synthetic data generation, or LLM-as-a-judge methods.

Education:Bachelor's or Master's degree in Computer Science or related field

(P25822_3552624)

#LI-Hybird

#LI-BB1