Principal AI Engineer — Agentic Systems Architect

Talentify; Santa Clara (CA); Hybrid; USD 176,000 - 293,000; Full time

Job summary

BMC Software in Santa Clara, CA, seeks a Principal AI Engineer to architect, build, and harden agentic AI systems powering our products. This is a hands-on, full-stack AI role spanning model lifecycle to agent orchestration.

This role will set technical direction, drive production-grade implementations, and collaborate across teams to deliver reliable, governable deployments with strong evaluation and observability.

Qualifications

• 8+ years of building and operating scalable, production-grade software with significant recent depth in AI/ML.
• Proven experience designing, building, and shipping agentic AI systems (autonomous/multi-step agentic workflows, multi-agent frameworks, and Generative AI copilots) in production.
• Hands-on experience building evaluation and experimentation frameworks for LLM/agentic systems (offline + online evals, LLM-as-a-judge, benchmarking, CI/CD gates, production monitoring).
• Strong agentic engineering skills, with modern agent/LLM frameworks (e.g., LangGraph, Google ADK, or comparable).
• Experience with RAG and retrieval (vector stores, grounding, citations) and familiarity with AgentOps/LLMOps and production monitoring.
• Solid software engineering fundamentals with productization tooling: Git, Docker, Kubernetes, CI/CD, and cloud (AWS / Azure / GCP).
• The judgment to operate in fast-moving, ambiguous environments and the communication skills to lead without authority.
• Experience with model customization (SFT, RLHF, DPO/GRPO), eval/observability platforms and bridging applied research and engineering is a nice to have.

Responsibilities

• Own the architecture for agentic systems end to end - reasoning and planning, tool/function calling, multi-agent coordination, routing, memory, and human-in-the-loop handoffs and define reusable patterns and reference implementations.
• Prototype new agentic capabilities quickly, then drive the ones that prove into production-grade systems; write production-quality code and set the engineering bar.
• Build and integrate across the stack: RAG and knowledge services, model/tool routing, prompt and context management, orchestration runtimes, and inference serving. Evaluate and adopt emerging models, frameworks, and protocols.
• Stand up the evaluation and experimentation strategy that gates what we ship, offline suites and golden datasets, regression tests in CI/CD, online evaluators on production traffic, calibrated LLM-as-a-judge graders, and A/B experiments to ensure safe and reliable deployments of agents in production.
• Define the metrics that matter (task success, grounded-ness, cost, safety, etc.) and define and build the tracing and observability to measure them across multi-turn interactions, closing the loop from error analysis to continuous improvement.
• Set technical direction across teams, mentor engineers, and translate complex architectures into clear guidance for partners and customers.

Skills

AI/ML Engineering

Agentic AI Systems

RAG and grounding

Cloud & Kubernetes

LangGraph / ADK

Model customization

Evaluation frameworks

Production-grade software

Tools

LangGraph

Google ADK

Vector stores

Git

Docker

Kubernetes

CI/CD

Job description

BMC Software in Santa Clara, CA, seeks a Principal AI Engineer to architect, build, and harden agentic AI systems powering our products. This is a hands-on, full-stack AI role spanning model lifecycle to agent orchestration.

This role will set technical direction, drive production-grade implementations, and collaborate across teams to deliver reliable, governable deployments with strong evaluation and observability.