Principal Architect - Autonomous Agent Safety & Control Systems

When you join Verizon

You want more out of a career. A place to share your ideas freely — even if they’re daring or different. Where the true you can learn, grow, and thrive. At Verizon, we power and empower how people live, work and play by connecting them to what brings them joy. We do what we love — driving innovation, creativity, and impact in the world. Our V Team is a community of people who anticipate, lead, and believe that listening is where learning begins. In crisis and in celebration, we come together — lifting our communities and building trust in how we show up, everywhere & always. Want in? Join the #VTeamLife.

Why Build AI at Verizon?

Most AI engineers are stuck building isolated chatbots, tweaking prompts, or fine-tuning APIs in low-stakes sandboxes with human babysitters in every loop. At Verizon, we're architecting the agentic nervous system powering hundreds of mission-critical internal applications, voice systems, and automated workflows executing millions of real decisions every day.

The real engineering thrill here isn't just giving agents autonomy—it's solving the hardest systems challenges in modern software: building the deterministic guardrails, sub-second circuit breakers, and continuous evaluation architectures that validate, observe, and optimize autonomous workflows at scale. You will combine advanced AI engineering skills with our governance frameworks to unleash the power of enterprise AI. With strategic partnerships spanning Google Cloud, Nvidia, AWS, and Anthropic, you won't just run experiments in a sandbox—you'll be the lead builder at the center of Verizon's most critical AI infrastructure.

What you’ll be doing...

As a Principal hands-on architectural engineer and technical lead within the Enterprise AI Control Plane team, you are the chief builder of the systems that make our AI faster, cheaper, and safer. You will architect the complex pipelines connecting Verizon’s AI developer platforms (AWS Bedrock, Atlas AI) with our centralized governance hubs (AEGIS, KAIROS). Your focus is deeply technical: designing enterprise-wide agentic evaluation systems, optimizing model inference, and deploying the zero-trust firewalls required for autonomous agents to run safely at a Fortune 20 scale.

In this role you will...

• Architect Agentic Evaluation: Design and code the core engineering systems for continuous AI agent validation, real-time observability, and automated evaluation metrics (e.g., LLM-as-a-judge).
• Drive Optimization & Fine-Tuning: Lead the fine-tuning of LLMs specifically to increase system throughput and aggressively reduce inference costs for agentic execution pipelines.
• Implement Reinforcement Learning: Utilize advanced RL techniques—including both harness tuning and model weights tuning—to optimize autonomous agent accuracy and decision reliability.
• Build Semantic Routing & Classification: Design and deploy lightweight classification models (e.g., JEV) to intercept, triage, and intelligently route traffic, minimizing expensive frontier LLM calls.
• Automate Governance: Architect and code automated CI/CD pipelines that inject Security and RIL (Responsible AI, Integrity, Legal) guardrails directly into the deployment lifecycle.
• Cross-Functional Technical Leadership: Serve as the technical anchor, interacting daily with AI Platform Engineers, DevOps/DataOps teams, and AI product developers to integrate their models seamlessly into the KAIROS control plane and AEGIS registry.
• Design IAM & Boundary Logic: Architect Identity & Access Management (IAM) and routing logic within the Agent Gateway to ensure autonomous workflows operate with strictly enforced, deterministic access boundaries.

What we’re looking for...

You've moved beyond standard RAG prototypes to solve the hard, unglamorous systems challenges of autonomous enterprise AI. You are a hands-on builder equally comfortable designing multi-agent execution frameworks, optimizing model weights, and building the sub-second circuit breakers that prevent autonomous systems from going rogue.

You will need to have...

• Bachelor’s degree or four or more years of work experience.
• Six or more years of relevant experience required, demonstrated through one or a combination of work and/or military experience, or specialized training.
• Four or more years of hands-on experience as an AI Engineer/Architect designing and deploying scalable, multi-agent frameworks leveraging interaction standards such as Model Context Protocol (MCP) and Agent-to-Agent (A2A) communications.
• Deep engineering experience in LLM operationalization, orchestration frameworks, model evaluation pipelines, fine-tuning, and automated safety testing.
• Experience optimizing LLM inference costs and throughput, utilizing classification models or semantic routing to reduce large-model dependencies.
• Experience in cloud architecture (AWS Bedrock, GCP) and CI/CD deployment automation.
• Experience defending agentic ecosystems against prompt injection, data exfiltration, tool abuse, and hallucinations in production environments.
• Experience shipping software into high-traffic, regulated enterprise environments where downtime or out-of-bounds agent actions carry severe real-world consequences.

Even better if you also have...

• Masters or PhD in Computer Science, AI/ML, or a related field.
• Six or more years of platform engineering, applied ML engineering, or AI science.
• Advanced programming skills in Python, Go, and Kubernetes.
• Hands-on expertise with advanced model optimization techniques (Reinforcement Learning, RLHF/RLAIF, LoRA, QLoRA).
• Experience building real-time telemetry, trace-level monitoring (OpenTelemetry, Langfuse, Arize), and automated fail-safes. (e.g., When an agent enters a recursive tool-calling loop or attempts an unauthorized action, your systems isolate the process and revert state in milliseconds).
• Knowledge of automated safety testing, red-teaming methodologies, and Agentic Zero-Trust IAM architectures for non-human identities.

If Verizon and this role sound like a fit for you, we encourage you to apply even if you don’t meet every “even better” qualification listed above.

Scheduled Weekly Hours

Equal Employment Opportunity 

Verizon is an equal opportunity employer. We evaluate qualified applicants without regard to veteran status, disability or other legally protected characteristics.

Benefits and Compensation

Our benefits are designed to help you move forward in your career, and in areas of your life outside of Verizon. From health and wellness benefit options including: medical, dental, vision, short and long term disability, basic life insurance, supplemental life insurance, AD&D insurance, identity theft protection, pet insurance and group home & auto insurance. We also offer a matched 401(k) savings plan, up to 8 company paid holidays per year and up to 6 personal days per year, paid parental leave, adoption assistance and tuition assistance, plus other incentives, we’ve got you covered with our award-winning total rewards package. Depending on the role, employees have the opportunity to receive compensation in the form of premium pay such as overtime, shift differential, holiday pay, allowances, etc. Newly hired employees receive up to 15 days of vacation per year, which grows with additional service. For part-timers, your coverage will vary as you may be eligible for some of these benefits depending on your individual circumstances.

The salary will vary depending on your location and confirmed job-related skills and experience. This is an incentive based position with the potential to earn more. For part-time roles, your compensation will be adjusted to reflect your hours.

The annual salary range for the location(s) listed on this job requisition based on a full-time schedule is: $120,500.00 - $231,000.00.

The annual salary range for the Illinois location(s) listed on this job requisition based on a full-time schedule is: $132,500.00 - $231,000.00.

Verizon will consider qualified applicants with criminal histories in a manner consistent with the requirements of the San Francisco Fair Chance Ordinance.