Software Engineer (agentic AI Applications)

Redfin Redfin · Seattle · Detroit, MI

Software Engineer to build and scale an AI-first product for mortgage bankers, focusing on agentic AI applications that analyze client communications and engagement signals to identify clients likely to convert. The role involves full-stack development, shipping product, and using AI as a core part of the software development lifecycle.

What you'd actually do

  1. Build and ship full-stack features from design through implementation, release, measurement, and iteration.
  2. Contribute to agentic AI capabilities such as agent orchestration, tool calling, retrieval-augmented workflows, and LLM-powered features.
  3. Help build evaluation, observability, and guardrail mechanisms for AI agents, ensuring reliability, safety, and measurable quality as agent behavior translates into product functionality.
  4. Implement scalable backend and frontend solutions using technologies such as C#, Python, Angular, AWS, and Kubernetes.
  5. Work with product, design, data, and senior engineers to turn requirements into reliable, well-tested solutions.

Skills

Required

  • 2+ years of professional software development experience
  • Experience building production software in languages such as C#, Java, Python, JavaScript, or TypeScript
  • Experience contributing to full-stack applications across frontend, backend, APIs, or data integrations
  • Familiarity with cloud platforms such as AWS, Azure, or Google Cloud
  • Solid grasp of software design fundamentals, testing, and writing maintainable code
  • Ability to communicate clearly and collaborate effectively across teams

Nice to have

  • Hands-on experience with LLM APIs, agentic AI frameworks (e.g., LangGraph, Semantic Kernel, AutoGen, CrewAI, or similar), or Model Context Protocol (MCP) tool integrations
  • Exposure to LLM application patterns such as retrieval-augmented generation (RAG), prompt engineering, or evaluating AI output quality
  • Hands-on usage of AI tools in your development workflow, such as AI coding agents, AI-assisted reviews, or prompt-driven development
  • Experience with AWS, Kubernetes, CI/CD, or automated testing

What the JD emphasized

  • AI-first product
  • agentic AI applications
  • ship real product
  • AI is not a side tool
  • core part of how you design, code, test, debug, review, and ship software
  • build evaluation, observability, and guardrail mechanisms for AI agents

Other signals

  • AI-first product
  • agentic AI applications
  • ship real product
  • AI is not a side tool
  • core part of how you design, code, test, debug, review, and ship software