AI Security Engineer

Summary

This role primarily focuses on the design, review of AI Security architecture and as well as the deployment and management of AI Solutions at Celestica. It includes (but not limited to) the secure deployment of AI products/services, administration and governance of Ai Models (Anthropic, OpenAI, Google Gemini, etc.), deployment and management of Ai Agents and Infrastructure supporting Ai use cases. This role will also provide governance and oversight into the consumption of Ai, providing input into aspects such as the creation of usage policies, forecasting and optimizing token consumption, model selection while maintaining stringent security controls.

This role reports into the Corporate IT Security leadership and will work within the IT function and the business to ensure that we meet and exceed the security requirements from Celestica customers. This role will also have to coordinate with external vendors and internal IT teams to ensure that the cybersecurity requirements are met and exceeded.

Detailed Description

Performs tasks such as, but not limited to, the following:

• Architects, designs, and secures Ai Solutions in multi-cloud environments, specifically focusing on Google Cloud Platform (GCP), Microsoft Azure, and AWS to enable secure deployment of enterprise AI solutions.
• Setup, maintenance and ongoing governance of various subscriptions to Ai Vendors such as Anthropic and OpenAI, (including AI Fin Ops)
• Architects, builds, and enforces security guardrails for AI Agents, establishing robust governance frameworks to monitor and control agentic behaviors.
• Participates in the security review of new applications with Ai Features.
• Leads the product selection or review for new cloud and AI security technologies, conducting Proof of Concepts (POCs) and technical evaluations
• Establishes and maintains effective relationships with process owners/sites to proactively assess business risks and develop risk mitigation.
• Design and Participation in the continued management and maintenance of AI security policies, Standards, guardrails and roadmaps.
• consumption of internal teams, namely Vertex Ai and Azure AI Foundry.
• Design AI Guardrails and rules to Secure AI Agents and data.
• Establishes and maintains effective relationships with process owners to proactively assess AI-related business risks, threats, and mitigation strategies.
• Designs and implements security controls for SaaS and PaaS AI integrations to ensure continuous data protection, DLP, and regulatory compliance.
• Communicates governance and compliance objectives, fostering a compliant and risk-aware security culture across internal engineering and product teams.

Knowledge/Skills/Competencies

• Operational knowledge and experience deploying and running solutions in cloud (Microsoft Azure, Google Cloud Platform and AWS). Includes knowledge of Cloud and Ai Fin Ops
• Network
• Demonstrated development experience including Coding or Scripting experience (ie. Python, Powershell, Java, .NET)
• Knowledge and proficiency Managing and Supporting solutions on Windows and Linux.
• Experience supporting IT Infrastructure such as Virtual Machines, Networking, DNS
• Knowledge and Experience in the deployment and management of Docker and Kubernetes images (including their hosting infrastructure)
• Knowledge and Experience in ML Pipeline Architecture. Concepts of Data Ingestion, Authentication and Data security, Model Registries, Serving and Hosting infrastructure including knowledge of how Models are trained
• Strong expertise in administering, configuring, and monitoring enterprise AI models including Claude, ChatGpt, and Google Gemini. Experience building governance frameworks for AI Agents, managing agentic permissions, tool execution, and runtime security controls.
• Must have strong knowledge of identity and Access Management, API Connectivity, Model Context Protocol (MCP), Data Security and API security
• Familiarity with Security Concepts such as:
  • • Adversarial examples and evasion attacks
    • Data poisoning and training-time attacks
    • Model extraction/stealing and inversion attacks
    • Membership inference attacks
    • Prompt injection (direct and indirect) and jailbreaking
    • Supply chain risks in pre-trained models, datasets, and third-party fine-tunes
    • Backdoors/trojans embedded in models
• Familiarity with Data Concepts such as (But Not limited to):
  • Data lineage and provenance tracking
  • Access controls over training data and embeddings
  • Deep understanding of AI vulnerabilities and experience in securing LLM’s within the Cloud environments.
• Knowledge and experience in IT Security Best Practices, IT Governance and Audit Procedures. Knowledge of common information security frameworks and IT controls frameworks, such as ISO/IEC 27001, ITIL, COBIT/COSO and ones from NIST.
• Excellent written and verbal communication skills, interpersonal and collaborative skills, and the ability to communicate security and risk-related concepts to technical and nontechnical audiences.
• Strong presentation skills for communicating architectural designs to executive leadership.
• Leadership skills, ability to motivate people.

Physical Demands

• Duties of this position are performed in a normal office environment.
• Duties may require extended periods of sitting and sustained visual concentration on a computer monitor or on numbers and other detailed data. Repetitive manual movements (e.g., data entry, using a computer mouse, using a calculator, etc.) are frequently required.
• Occasional overnight travel is required.

Typical Experience

• 10+ years previous experience in a similar role/industry.

• Extensive hands-on experience designing and securing Google Cloud, Azure, and Google Workspace environments.

Typical Education

• Bachelors Degree.
• CISSP, CCSP (Certified Cloud Security Professional), or relevant cloud vendor certifications are a plus.

Notes

This job description is not intended to be an exhaustive list of all duties and responsibilities of the position. Employees are held accountable for all duties of the job. Job duties and the % of time identified for any function are subject to change at any time.