> Markdown version of [/jobs/ext/4769404-principal-full-stack-engineer-ai-products](https://www.wearedevelopers.com/jobs/ext/4769404-principal-full-stack-engineer-ai-products). Every page supports `.md` or `Accept: text/markdown`. Links point to the HTML versions so they work for humans too. Agent guide: [/agents.md](https://www.wearedevelopers.com/agents.md). --- # Principal Full Stack Engineer, AI Products - **Company:** The Baker - **Location:** Seattle, WA, United States - **Experience:** Expert - **Salary:** $200,000 - $350,000 - **Contract:** Permanent contract - **Skills:** Application Programming Interfaces (APIs), Artificial Intelligence, Audit Trail, Microsoft Azure, Data Governance, Python (Programming Language), TypeScript, ReactJS, Large Language Models, AI Coding Agents, Backend, Agentic-AI, Data Layers, Databricks - **Published:** November 18, 2026 - **Apply:** https://www.disabledperson.com/jobs/76556454-principal-full-stack-engineer-ai-products ## About the Role * You're a hands-on technologist who has built things people actually use, ideally end to end, from the first conversation with users through production and after. * You're naturally curious, a strong problem solver, and a clear communicator. * You can take an objective, form an opinion, and move things forward on your own, and bring others with you, even when the path is unclear. What you bring * 9+ years building and running production software, including complex products you built end to end. * You use AI coding tools extensively, and you have opinions about what works. * Deep experience in a backend language, and the judgment to guide architecture across web, services, APIs, and data. * You've designed systems where access control, data governance, and audit trails were first-class requirements. * You've shipped LLM or agent systems to production, and you have a clear view on evaluation, retrieval, and tool integration. * You explain hard tradeoffs clearly and guide several engineers through them. * You've raised the bar for a team through reusable solutions, good decisions, and mentoring. * A technical degree or equivalent experience. Nice to have * Shared standards or platforms that several teams adopted. * Databricks or lakehouse architectures, graph and vector databases, MCP, or LLM tracing and evaluation tools. * Evaluating third-party AI vendors against an internal standard, or working with enterprise identity and regulated workflows. ## Description We're building AI agents that do real work for Baker Tilly's tax, assurance, and advisory teams. They read client files, find risks and opportunities, draft financial statement disclosures, and show their sources, and a practitioner always makes the final call. The data is some of the most sensitive in the business, so we build security in from the start., As a Principal Full Stack Engineer, you'll be the technical anchor for this team: hands-on, setting the standards for how our agents are built and trusted, and growing the engineers around you. It's an individual-contributor role. You lead with code, judgment, and influence. What you'll work with * Years of client engagement history, tax returns, and financial records, at the scale of a major firm. * Tax, audit, and advisory practitioners who help define what good looks like and review what the agents produce. * Real users inside the firm who give direct feedback on what you build. * Close collaboration with Ode with Anthropic, an AI engineering company the firm announced an initiative with in July 2026. * A mix of building and buying. You'll help decide which is which, and everything we ship meets the same standards for security, tracing, and testing. * Our stack: Python, TypeScript and React, Databricks, graph and vector databases, MCP, OpenTelemetry, and Braintrust, on Azure. What you'll do * Set technical direction for a growing set of agents and the data layer underneath them, with secure, well-defined interfaces between them. * Design how agents handle sensitive data: access that follows the person asking, enforced when data is retrieved, with audit records and human approval for what agents create. * Set the bar every agent must meet, whether we build, buy, or integrate it: security, tracing, testing, production readiness. Own how we decide between build and buy. * Write code. Build the high-leverage pieces and take on the hardest production problems yourself. * Make testing a core discipline: expert-labeled ground truth, regression checks, and honest numbers on accuracy, cost, and risk. * Work with product, security, data, legal, and practitioner leaders to settle tradeoffs, and mentor senior engineers through design reviews. * Evaluate models and tools with evidence, and keep humans in the loop where the stakes call for it. * Shape how we ship with AI as we move toward a code factory: the tools, workflows, and standards that keep AI-written code safe in production.