AI Platform Security Engineer
at Kai · 11-50 employees
- Location
- San Jose, CA
- Posted
- 20d ago
at Kai · 11-50 employees
Kai is an artificial intelligence-powered cybersecurity startup.
We're looking for a AI Platform Security Engineer to drive the security of the Azure infrastructure that powers the Kai AI-native cybersecurity product. This role centers on the security of the cloud foundation, data platform, AI/ML infrastructure, and internal developer platform that the product depends on. This is a deeply technical, infrastructure-focused role. You'll work closely with Platform Engineering, DevOps, Data Engineering, and AI/MLOps teams to ensure that the systems, pipelines, and environments underpinning our product are designed, built, and operated securely. What You'll Do... Cloud Infrastructure Security Own the end-to-end security infrastructure architecture of our Azure environment, including landing zone design, management group and subscription structure, network topology, and resource governance. Enforce and continuously improve guardrails using Azure Policy, Cloud security posture management (CSPM), and infrastructure-as-code (IaC) security scanning (Checkov, tfsec, or equivalent). Manage and mature the Azure network security model: hub-and-spoke topology, NSG and Azure Firewall rule governance, Private Endpoints, and DDoS protection controls. Lead cloud infrastructure security posture reviews, drive down misconfigurations, and own the organization's Secure Score improvement roadmap. Maintain and harden Azure landing zones, ensuring new workloads are provisioned into a secure-by-default environment. Identity, Access, and Secrets Management Drive the organization's cloud identity and access management strategy, including Entra ID tenant configuration, Privileged Identity Management (PIM), Conditional Access policies, and workload identity (managed identities, federated credentials, service principals). Enforce least-privilege IAM across all Azure subscriptions and resources; conduct regular access reviews and entitlement hygiene campaigns. Architect and operate the enterprise secrets management program using Azure Key Vault with HSM-backed keys, including key rotation automation, certificate lifecycle management, and developer-facing secrets injection patterns. Define and enforce policies for human and non-human identities across CI/CD systems, internal tooling, and AI/ML workloads. Kubernetes and Container Platform Security Secure the Azure Kubernetes Service (AKS) platform: cluster hardening, node pool configuration, admission control (OPA/Gatekeeper, Kyverno), runtime security, and network policy enforcement. Own container security standards: base image governance, image signing and provenance (Notary, Cosign), container registry security (Azure Container Registry), and vulnerability scanning integration in the build pipeline. Maintain and improve Pod Security Standards, workload identity binding (Azure Workload Identity), and namespace-level security isolation. Collaborate with Platform Engineering on the internal developer platform (IDP) to ensure that developer self-service pathways are built with security guardrails as first-class controls. AI and Data Platform Security Secure the data and AI/ML infrastructure layer. Define and enforce data security controls including storage encryption (CMK), data classification enforcement, network isolation for data services, and access boundary policies between training, staging, and production AI environments. Establish security controls for AI/ML pipelines: training data provenance and integrity, model artifact signing, inference endpoint hardening, and isolation of multi-tenant AI workloads. Work with Data Engineering and MLOps teams to ensure AI infrastructure changes go through security review and that data access patterns are auditable and compliant. Detection, Response, and Vulnerability Management Own the cloud-native detection and monitoring stack Develop and maintain detection rules and analytic content tuned to cloud infrastructure and AI platform threats (e.g., credential abuse, lateral movement, data exfiltration from AI workloads). Lead the infrastructure vulnerability management program: agent-based and agentless scanning across Azure VMs, AKS nodes, and container images; SLA-based remediation tracking; and patch compliance reporting. Own cloud incident response runbooks for infrastructure-layer security events and serve as the technical lead for cloud-scoped security incidents. Security Automation and Platform Hardening Build and maintain policy-as-code frameworks that enforce security standards across IaC templates (Terraform, Bicep) before resources are provisioned. Develop internal security automation for drift detection, misconfiguration remediation, and continuous compliance validation against CIS Azure Foundations Benchmark and equivalent baselines. Partner with DevOps and Platform Engineering to embed security gates into infrastructure CI/CD pipelines, ensuring that insecure infrastructure changes cannot reach production. Maintain the platform security baseline documentation and runbooks, enabling the broader engineering organization to build a well-understood, secure foundation. What We're Looking For Required An ownership mentality that places the wellbeing of the company, our customers, and teammates at the forefront of everything that the role does. Ability to thrive in a high-paced, high-growth startup environment. 6+ years of experience in cloud security, infrastructure security, or platform security engineering, with at least 3 years working deeply in Microsoft Azure . Expert-level knowledge of Azure security services: Entra ID, Key Vault, Azure Firewall, Azure Policy, and Private Networking. Strong hands-on experience with Kubernetes security and AKS platform operations, including admission controllers, runtime security, and workload identity. Demonstrated experience securing data platforms and AI/ML infrastructure (data lakes, blob storage, model training environments, inference endpoints). Proficiency with infrastructure-as-code tools (Terraform and/or Bicep) and IaC security scanning.