SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Jerry.ai is a profitable, Series C, YC-backed fintech startup with 5M+ customers and $240M in funding, building an AI-powered advisor to manage physical assets. Starting with car insurance in 2019, the company is now expanding into home, motorcycle, RV, and boat insurance verticals. The company has achieved significant automation milestones: 60% of policies are sold automatically without human intervention, and 90% of inbound customer inquiries are resolved via AI.
The AI & Automation team owns the critical infrastructure that automates the end-to-end insurance quoting experience and orchestrates complex, multi-step transactions with 100+ insurance carrier systems. Before off-the-shelf AI agents existed, Jerry built proprietary infrastructure to scale automated insurance purchases and customer support. Today, 78% of inbound inquiries are resolved via AI with 83% positive experience ratings.
As Software Architect, you will lead the design and implementation of the next-generation AI system. Your responsibilities include: setting the technical bar for chatbot systems and guiding the prompt engineering team; building new automations for document validation (driver's licenses, proof of insurance), record disputes, and email/callout reduction; evaluating emerging AI technologies (voicebot, computer use agents) and making investment decisions; and partnering with OpenAI to evaluate and deploy next-generation voice models and workflow automation.
You will own outcomes end-to-end, from architecture decisions through production deployment. The role requires hands-on depth with AI systems, including production experience with LLMs covering evals, regressions, latency optimization, cost management, and failure handling. The tech stack includes TypeScript (Nest.js, React), GraphQL, AWS (ECS/EKS, Lambda), and LLM models from OpenAI, Anthropic Claude, and Google Gemini.
Ideal candidates have 5+ years of software engineering experience, preferably at early-stage startups where they owned outcomes end-to-end. You should be a builder comfortable working without detailed specs, with real hands-on depth in AI—having built with new models within days of release and able to articulate why one architecture beats another beyond demos.
