Role:
We are seeking a Staff Platform Engineer (AI Systems) to join our Core Team. Strike is already a high-velocity, cloud-native engineering organization. We run an immutable infrastructure on GKE, we enforce "Shift-Left" security, and we deploy constantly. We are full on integrating frontier AI capabilities directly into our ecosystem - not as a gimmick, but as a force multiplier for our team.
This is a Systems Engineering role. You will not be fine-tuning models or building chatbots for fun. You will be building the secure connective tissue that allows agentic systems to interact with our real-world infrastructure. You will define how agents authenticate, how they access tools, and - crucially - when they should be trusted.
Who this is NOT: To save your time: This is NOT a Data Science or MLOps role. We are not looking for someone to train custom models in PyTorch or manage Kubeflow clusters. We are looking for a systems builder to create the agentic harness and tools that empower our team.
What You Will Do
- Drive the Vision: Blend your ability to execute with your ability to strategize. You will take the insight you gather working with teams, translate that into a vision, and influence organizational alignment on overall AI strategy.
- Build the Internal AI Platform: We have already operationalized tools like Copilot Agents, Gemini, Claude, Agentic CLIs, n8n workflows, and local agent runtimes. Your job is to build the unified, secure layer that connects these tools to our core platforms (e.g., building internal MCP servers).
- Create IDE Agents: Help us fine-tune and deploy IDE agents that accelerate coding for the entire engineering organization.
- Develop Agentic Workflows: Create orchestration layers and secure runtime environments for LLM tool-use.
- Bridge Information Silos: Build out the enterprise knowledge base to connect disparate data sources for agentic reasoning.
- Define Tiered Autonomy: You will design the security patterns for different levels of agent independence - from "Advisory" (read-only) to "Autonomous" (action-taking). You will implement the "Verifiable State" and "Revert" mechanisms that make Tier 3 autonomy safe.
- Secure the Frontier: Design the auth patterns for agentic access. You will solve hard problems around authentication, authorization, and "human-in-the-loop" safeguards, ensuring we move fast without breaking our "Zero Trust" principles.
- Pragmatic Building: You will be the voice of reason. You know when to use an LLM, and when to just write a script. You will help teams identify "Units of Toil" and target them with the right level of abstraction.
- Force Multiply the Team: Embed with teams to unblock their workflows. You will serve as the expert on frontier tools, helping engineers and the broader team optimize their setups and adopt "AI Native" workflows.
What We Are Looking For
- Full Stack Systems DNA: You are a polyglot engineer. You understand the entire stack - from frontend interfaces to backend distributed systems and cloud infrastructure.
- Frontier AI Proficiency: You are fluent in the current landscape of AI enablement. You have built implementations using patterns like MCP, RAG, function-calling, and evals. You understand techniques like AI as a Judge, Agent Self-Improvement, and how to balance deterministic systems with non-deterministic outputs.
- Security-First Mindset: You understand the risks of deploying LLMs in a production environment. You are familiar with concepts like Shift-Left security, least-privilege access, and robust secrets management.
- Organizational Influence: You have excellent communication, collaboration, and influencing skills. You can explain complex AI concepts to non-technical stakeholders and drive consensus on architectural decisions.
- Pragmatism: You treat AI as a component, not magic. You are skeptical of hype and focused on measurable utility. You understand the need to balance the cognitive load of LLM output without sacrificing the efficiency gains.
- Ownership: You are an owner at heart. The desire to take ownership of your domain is innate and automatic to you. You operate autonomously and see things through to completion - every time.
Our Environment (Context)
- Infrastructure: Fully Cloud-Native on GCP, Infrastructure as Code
- Orchestration: Google Kubernetes Engine (GKE) with Helm & ArgoCD.
- Languages: C#/.NET (Core Services), Python (Data/Scripts), TypeScript (Web).
- Security: Comprehensive enterprise secrets management and governance tooling.
- AI Stack: We use a mixture of tools and models; being flexible and ready for change is the default.