AI Engineering Platform Architect

Socket.dev; Salt Lake City (UT); On-site; USD 150,000 - 230,000; Full time

Job summary

Onebrief is seeking an Outcome Engineer to join our AI Engineering Platform team. You’ll design and build scalable agentic workflows, moving beyond prompts to robust, stateful systems that automate complex decision making.

You’ll collaborate with engineers, researchers, and customers to ship rapid prototypes, prove concepts in production, and implement evaluation frameworks that formalize agent performance.

Qualifications

• Experience designing multi-agent architectures and orchestration layers.
• Hands-on experience building agentic AI platforms or similar systems.
• Strong software engineering fundamentals and scalable design skills.
• Familiarity with distributed systems and microservices patterns.
• Proficiency in Node.js and TypeScript stack.
• Experience with vector databases and semantic tooling.
• Knowledge of observability and monitoring practices.

Responsibilities

• Architect Multi-Agent Systems to enable agents to collaborate, debate, and execute complex tasks.
• Implement Automated Governance with policy engines and constraints.
• Engineer Context and Memory using knowledge graphs and vector stores.
• Build Rigorous Evaluation Frameworks with deterministic validation and metrics.
• Design Self-Healing Loops by linking observability to agent actions.
• Prototype and Deploy new agentic tools to production for internal and external use.
• Experiment with new models and orchestration techniques daily.

Skills

Multi-Agent Systems

Agentic AI

System Architecture

Distributed Systems

Node.js/TypeScript

Vector Databases

Observability

Job description

Onebrief is seeking an Outcome Engineer to join our AI Engineering Platform team. You’ll design and build scalable agentic workflows, moving beyond prompts to robust, stateful systems that automate complex decision making.

You’ll collaborate with engineers, researchers, and customers to ship rapid prototypes, prove concepts in production, and implement evaluation frameworks that formalize agent performance.