Senior Applied AI Engineer - Agentic Systems

Senior Applied AI Engineer Agentic Systems

Location: Mountain View, CA

Job Description
Agentic Feature Development & Full Stack Delivery
Design, build, and ship agentic features directly within EAS - autonomous workflow agents, multi-step task orchestration, tool-calling loops, and human-in-the-loop interaction patterns
Own agentic features end to end - from architecture and implementation through testing, hardening, and production deployment
Identify high-value automation opportunities within EAS workflows and translate them into well-scoped, shippable features
Integrate new agentic capabilities cleanly into an existing product codebase without disrupting existing functionality
Own the full stack of agentic feature delivery - backend orchestration, API integration, and front-end surfaces that expose agent capabilities to enterprise users
Build RAG pipelines over structured and unstructured data to power intelligent retrieval, decision support, and workflow automation within EAS
Build memory and state management systems that allow agents to maintain context across multi-step, long-running workflows

Agentic AI - Core Requirement
Build production-grade agentic systems with the reliability, observability, and failure handling that enterprise software demands
Design evaluation harnesses to continuously test agent accuracy, behavioral consistency, and edge case handling
Build guardrails, fallback logic, and escalation patterns that ensure agents degrade gracefully and keep users in control
Instrument agentic features with logging, tracing, and monitoring to observe agent behavior in production and iterate with confidence
Participate in architecture and design reviews - contributing agentic expertise and maintaining quality standards across short delivery cycles
Contribute to shared agentic patterns and reusable components that raise the capability baseline for the broader EAS engineering team
Define and implement evaluation frameworks to measure agent accuracy, task completion, and behavioral consistency across diverse inputs and edge cases
Experience with both automated eval pipelines (unit-level tool call testing, end-to-end trace evaluation) and human-in-the-loop review workflows for validating agent outputs in production

Required Experience
Demonstrated hands-on experience building agentic AI capabilities inside a product - multi-step orchestration, tool-calling agents, memory systems, and human-in-the-loop flows used by real users in production
Deep familiarity with agent frameworks - LangGraph, Anthropic SDK, OpenAI Agents SDK, CrewAI, AutoGen, or similar - applied in product feature delivery, not research
Strong understanding of agentic design patterns: planning loops, tool registries, context window management, agent state machines, and failure handling
Experience building and integrating RAG pipelines into product workflows
Experience building production guardrails and evaluation frameworks for agentic features
Strong full-stack engineering skills with production experience in Python and/or TypeScript
5+ years of full-stack software engineering with a strong shipping record
1+ years of hands-on experience building agentic AI features in production products

Preferred
Experience integrating agentic capabilities into SaaS or fintech products at scale
Familiarity with Intuit's developer platform or QuickBooks APIs
Exposure to regulated or high-accuracy domains where agent reliability and auditability are non-negotiable