AI Engineer at Withshepherd

You will join a focused team of four building the world’s first fully autonomous underwriting system for high-hazard industries. By architecting agentic workflows and LLM pipelines, you’ll transform complex, unstructured insurance documents into real-time pricing decisions, eliminating weeks of manual delays for critical infrastructure projects like data centers and renewable energy assets.

• Architect and deploy agentic workflows that chain LLM capabilities to perform autonomous insurance underwriting decisions.
• Design and optimize RAG pipelines to turn messy, unstructured document submissions into 100% accurate structured data.
• Collaborate directly with the CTO and insurance experts to iterate on product strategy and high-impact AI implementations across the stack.

• 5+ years of full-stack experience (TypeScript/Node.js/React) with a track record of delivering complex, production-ready products.
• Deep hands-on experience with LLM APIs, prompt engineering, and building agentic systems or AI-powered extraction pipelines.
• An entrepreneurial mindset that thrives in high-ambiguity environments and takes end-to-end ownership from conception through to release.

• Ship daily in a high-growth environment that treats AI (including Claude Code) as a core part of the internal engineering workflow.
• Work at a $60M+ Series B startup backed by world-class investors like Spark Capital, Costanoa, and one of the world’s largest insurers.
• Build zero-to-one agentic systems that enable ‘email in, price out’ automation for a $700B+ industry that traditionally relies on static legacy processes.