Software Engineering SMTS

Salesforce•San Francisco, CA

4d•Hybrid

About The Position

Salesforce AI Labs is building a product that learns from human demonstrations and turns complex workflows into repeatable automation across enterprise applications. The browser gives AI access to an enormous world of legacy enterprise software. Turning that access into dependable action requires understanding how applications actually behave: dynamic interfaces, nested frames, custom controls, and interactions that vary from one execution to the next. The primary focus will be reliable browser automation, with work across the stack to make those capabilities usable and diagnosable by customers. Understanding how browser behavior connects to frontend state and backend services is central to making the whole system work. This role involves building directly with customers, shaping the product's capabilities, and owning capabilities from initial design through production use.

Requirements

Independently owned a significant web capability or system through design, deployment, production failures, and subsequent improvements.
Strong JavaScript or TypeScript skills and a deep understanding of the DOM, browser events, network behavior, and asynchronous execution.
Experience building browser automation with Playwright or comparable tools.
Ability to trace difficult failures across browser behavior, frontend state, and backend services.
Ability to design tests that distinguish reliable behavior from something that happened to work once.
Desire to work directly with customers, investigate unfamiliar applications, and turn learnings into product improvements.
Thoughtful use of AI development tools and responsibility for the correctness, security, and maintainability of shipped code.

Nice To Haves

Browser extensions, accessibility trees, resilient element targeting, or recording and replay systems.
LLM-powered applications or computer-use agents.
Building an early-stage product or working in a customer-facing engineering role.

Responsibilities

Drive Browser Automation & Architecture: Build browser-extension capabilities that capture user interactions and the context needed to reproduce workflows reliably.
Develop automation for complex applications, including dynamic pages, nested frames, custom controls, and asynchronous state changes.
Improve how agents understand page structure, identify controls, execute actions, and recover when interfaces change.
Create debugging and replay tools that make automation failures easier to understand and resolve.
Extend the product experience so customers can record workflows, monitor execution, and inspect results.
Build Directly With Customers: Observe how customers use their applications and identify the interactions and state changes automation needs to capture.
Investigate failures in customer environments, tracing problems through browser behavior, application state, and execution services.
Turn application-specific discoveries into reusable recording, targeting, and recovery capabilities.
Help customers record and troubleshoot workflows independently, using their experience to improve the product.
Shape Technical Direction & Grow the Team: Own architectural decisions for browser recording and execution, including recording fidelity, element targeting, and browser state.
Partner with AI engineers to define the observations and actions agents need to operate applications reliably.
Mentor peers through design discussions and code reviews, and participate in hiring.