Senior Full-Stack AI Engineer: Enterprise Agent Workflows

Socket.dev; Austin (TX); On-site; USD 140,000 - 190,000; Full time

Job summary

Light & Wonder in Austin, TX, is seeking a senior software engineer to own complex agentic workflows from UI to model execution. You will navigate a React frontend, a C#/.NET orchestration layer, and tool integrations to deliver resilient, enterprise‑scale AI agents.

You should have broad technical depth across async patterns, API design, and Azure primitives, with hands‑on experience in production systems using LLMs.

Qualifications

• 6+ years of production software engineering across frontend and backend.
• Deep proficiency in C#/.NET, TypeScript, and React.
• Experience shipping production systems with LLMs and multi‑step orchestration.

Responsibilities

• Architect end‑to‑end agent workflows spanning UI, orchestration, and tool integrations.
• Build embedded agent interfaces using AG‑UI and CopilotKit for refined UX.
• Implement orchestration with .NET Agent Framework and handle multi‑agent handoffs.
• Integrate with low‑code platforms like N8N to accelerate delivery while staying reliable.
• Develop robust evals and testing for production‑grade agent behavior.
• Collaborate with Data Engineering and Platform Engineering on data surfaces and deployment standards.
• Contribute to reusable platform primitives and code review practices.

Skills

C#/.NET

TypeScript

React

LLM integration

Async patterns

API design

Education

Bachelor's degree in CS

Tools

Agent Framework

AG-UI

CopilotKit

N8N

Azure

Snowflake

Databricks

Foundry Observability

Job description

Light & Wonder in Austin, TX, is seeking a senior software engineer to own complex agentic workflows from UI to model execution. You will navigate a React frontend, a C#/.NET orchestration layer, and tool integrations to deliver resilient, enterprise‑scale AI agents.

You should have broad technical depth across async patterns, API design, and Azure primitives, with hands‑on experience in production systems using LLMs.