Sr. Software Engineer, Agents

Globality•Palo Alto, CA

5d•Hybrid

About The Position

Globality is hiring a Senior Software Engineer to build the agentic platform at the core of its autonomous sourcing product. The role owns features end to end across React and TypeScript interfaces, backend services and APIs, and the orchestration layer where agents plan, call tools, and take action. The engineer will collaborate closely with product and design, influence architecture, and help shape how the team builds with coding agents. Agentic coding tools are intended to be the primary development workflow: the engineer will use them to plan, implement, test, and review work, applying judgment to ensure reliable production software. The position also involves developing the context, conventions, and guardrails that help coding agents work effectively in the codebase. The role is based in Palo Alto, with employees working in the office four days per week and remotely or wherever they are most effective one day per week. The anticipated annual base salary range is $150,000–$250,000, with actual compensation varying based on factors such as experience, skills, and location.

Requirements

At least 5 years of professional software engineering experience across the stack.
Hands-on experience shipping production software with agentic coding tools, including delegating multi-step tasks, running agents in parallel, and reviewing and correcting their output.
A clear point of view on working with coding agents, including structuring context and instructions, scoping tasks, verifying results, and deciding when to take over manually.
Experience building LLM-powered or agentic features in production, such as tool calling, retrieval, structured outputs, streaming, or multi-step workflows.
Solid backend fundamentals, including REST or GraphQL API design, relational data modeling, queues and background jobs, and asynchronous and event-driven patterns.
Experience delivering real-time data to browsers over WebSockets or server-sent events, including partial, streaming, and failure states.
Working knowledge of cloud infrastructure, containers, and CI/CD, and the discipline to own software in production.
Strong testing and code review habits, including when much of the first draft is produced by an agent.
Clear written and verbal communication, including the ability to explain technical tradeoffs to non-engineering stakeholders.

Responsibilities

Design, build, and ship full-stack features for an agentic sourcing platform, including React and TypeScript interfaces, backend services, data models, and APIs.
Build agent orchestration, including planning and tool-calling loops, multi-agent workflows, long-running tasks, retries, and human approval checkpoints.
Create human-agent collaboration interfaces with streaming output, live progress, reasoning and evidence views, confidence indicators, and controls to pause, redirect, or override agents.
Use coding agents as a primary workflow, running tasks in parallel, directing and reviewing their work, and ensuring the results meet production standards.
Build project instructions, reusable skills, test harnesses, and review checklists that make coding agents effective in the codebase.
Integrate LLMs and external tools through APIs and protocols such as the Model Context Protocol, managing prompts, context, cost, and latency.
Build evaluation, tracing, and observability for agent behavior so the team can measure quality, replay runs, and improve results.
Design for security, privacy, and auditability in line with SOC 2 Type 2, ISO 27001:2022, GDPR, and CCPA requirements, including permissions, data boundaries, and logging for autonomous actions.
Write unit, integration, and end-to-end tests, and review code—including agent-written code—for reliability, security, and maintainability.
Own production services, including deployment, monitoring, and performance tuning for real-time, data-intensive workloads.
Participate in sprint planning, architecture discussions, and technical direction alongside senior engineers.