Senior AI Engineer (Agentic Systems, Production)

Ozmo; United States; Remote; USD 162,000 - 198,000; Full time

Benefits offered by this job

Medical insurance

PTO

401k

Maternity leave

Pawternity leave

Sabbatical

Remote work flexibility

Job summary

You will lead the development of MCP servers, retrieval layers, and governance patterns while balancing latency, cost, and reliability. This fully remote role offers a path to impact large enterprise customers and mentor growing engineers.

Qualifications

• 7+ years of software engineering experience, with substantial experience building and operating production SaaS.
• 2+ years of hands-on production LLM/agent experience, with demonstrated systems shipped to real users.
• Proven track record of taking AI products from prototype through pilot and into production, ideally in customer-facing environments.
• Practical experience using AI coding agents such as Claude Code, Codex, Cursor, Gemini CLI, or similar tools as part of your day‑to‑day development.
• Deep hands-on experience building agentic systems, including tool use, orchestration, planning, multi‑agent workflows, and human‑in‑the‑loop systems.
• Strong understanding of AI infrastructure, particularly MCP, agent gateways, skills, memory, evaluation, observability, and interoperability.
• Strong knowledge of RAG and knowledge systems, including hybrid retrieval, vector + graph databases, metadata, taxonomy/ontology, and retrieval evaluation.
• Strong enterprise engineering fundamentals, including multi-tenancy, security, authorization, auditability, reliability, and distributed systems.
• Production cloud-native experience, including Kubernetes, CI/CD, infrastructure‑as‑code, Docker, and AWS/Azure/GCP.
• Expert‑level Python and strong software architecture skills, including API design, data modeling, testing, performance, and distributed systems.
• Exceptional technical judgment and communication, with the ability to explain non-deterministic AI behavior, make pragmatic tradeoffs, mentor engineers, and take ownership of outcomes.

Responsibilities

• Build and ship production-grade agentic AI products that solve real, mission-critical enterprise problems.
• Own the full product lifecycle from discovery and prototyping through customer pilots, GA, and ongoing iteration.
• Design agentic architectures with LangGraph, LangChain, CrewAI, PydanticAI, or equivalent: planning, decomposition, tool use, multi-step workflows, memory, state management, self-correction, and human-in-the-loop patterns.
• Build the agent infrastructure layer including MCP servers/tools, skills, agent gateways, and agent-to-agent communication.
• Build the knowledge and retrieval layer including RAG, hybrid graph/vector retrieval, taxonomy, ontology, metadata, provenance, and content lifecycle management.
• Design for enterprise-grade security and trust, including authorization, tenant isolation, auditability, guardrails, prompt-injection protection, and controlled tool access.
• Build rigorous evaluation and observability systems that measure task success, quality, latency, cost, drift, and reliability. Use them to catch regressions before release and monitor performance in production.
• Make sound architectural tradeoffs about when to use agents versus deterministic software, balancing accuracy, reliability, latency, complexity, and cost.
• Own systems in production end-to-end, including deployment, monitoring, incident response, performance, reliability, and cost.
• Create reusable patterns and tooling, mentor other engineers, and help establish how Ozmo builds with AI.

Skills

Python

LLM/Agent experience

Production SaaS

Kubernetes

Cloud-native

Security & auditability

Observability

Distributed systems

API design

Mentoring

Ownership

Tools

LangGraph

LangChain

CrewAI

PydanticAI

Job description

You will lead the development of MCP servers, retrieval layers, and governance patterns while balancing latency, cost, and reliability. This fully remote role offers a path to impact large enterprise customers and mentor growing engineers.