Software Engineer, MTS - Agent Studio & Observability UI

Salesforce•San Francisco, CA

2d•$117,200 - $194,200•Onsite

About The Position

Join the Agentforce Optimization Agent team at Salesforce, where we're building the first AI-native observability platform for agentic systems. Based in San Francisco, you'll help shape how customers see, understand, and improve their AI agents — crafting the interfaces that turn complex agent session data into clear, actionable insight directly inside Agent Studio.

Requirements

Strong, hands-on experience building production frontend applications in JavaScript/TypeScript and React, with a track record of shipping performant, accessible UI at scale.
Demonstrated, genuine AI-first approach to engineering — using AI tools to move faster, build fluency across the stack, and contribute well beyond your core specialty.
Advanced prompt engineering skills: the ability to write precise, structured prompts and cultivate the system context that makes AI outputs reliable, secure, and production-ready.
A related technical degree required.

Nice To Haves

Experience integrating frontend systems with real-time data sources, streaming APIs, or analytics/observability platforms.
Experience working within a large-scale platform or multi-cloud environment and understanding the constraints that come with it (security, compliance, multi-tenancy).

Responsibilities

Own the UI surface for agent observability: Build and ship production-grade React components and views that surface optimization insights, session analysis, and performance signals inside Agent Studio's Sessions and Details views — making complex AI data legible and actionable for developers.
Drive the Agent Studio integration: Lead frontend development across the Agent Developer Lifecycle (ADLC), integrating observability tooling into Salesforce Core and ensuring your components work reliably across multi-tenant, multi-cloud environments.
Solve real-world UI performance challenges: Tackle rendering and data-fetching performance at scale — designing frontend systems that handle high-volume agent session data with low latency and a seamless developer experience.
Build and ship using AI as a core workflow tool: Push the boundaries of AI development tools (Claude Code, GitHub Copilot, Cursor) to deliver secure, optimized, production-grade code — and contribute to the shared system context that keeps AI outputs reliable.