> Markdown version of [/jobs/ext/4604106-senior-software-engineer](https://www.wearedevelopers.com/jobs/ext/4604106-senior-software-engineer). 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). --- # Senior Software Engineer - **Company:** Pearson - **Location:** Durham, NC, United States - **Experience:** Expert - **Salary:** $95,000.0 - $160,000.0 - **Contract:** Permanent contract - **Skills:** LangGraph Framework, Artificial Intelligence, Information Engineering, Data Governance, Cursor, Python (Programming Language), Node.Js, OpenAI, Next.js, TypeScript, Microsoft Agent Framework, ReactJS, Claude Code, Multi-Agent Systems, Agentic-AI, Data Layers, AgentOps, CrewAI, Claude, Human in the Loop - **Published:** November 6, 2026 - **Apply:** https://dejobs.org/x/x/AE2ED786D3B24614AF923702AD9F25BB/job/ ## About the Role * 5+ years building and shipping production software, including recent hands-on experience designing, deploying and operating agentic applications in production. * Strong full-stack engineering across the modern web stack - TypeScript, React, Next.js, Node and Python - with sound architecture and system-design judgment. * Hands-on experience with agentic frameworks and orchestration - LangChain and LangGraph, Deep Agents, CrewAI, Google's Agent Development Kit, the Microsoft Agent Framework, the OpenAI or Claude Agent SDKs, the Vercel AI SDK or equivalents - and a clear view of the trade-offs between them. * Experience building the agent interface layer: generative and streaming UI with AG-UI, A2UI, CopilotKit or similar, and human-in-the-loop interaction patterns. * Multi-agent systems in production: tool-using agents via MCP, agent-to-agent coordination, and the evals and guardrails that keep them reliable. * Practical experience with voice and multimodal AI - speech-to-speech models, realtime audio and vision - and judgment on where they add value in a product. * A trace- and eval-driven approach: you instrument agents, read the traces, and design evals and guardrails from what the evidence shows. * Fluent daily use of agentic engineering tools such as Claude Code, Cursor, Codex or Antigravity, with the engineering judgment and review discipline to match. * Customer-facing delivery skill: you can run discovery with a business team, scope tightly, and communicate credibly with both engineers and senior stakeholders. * Enterprise delivery experience: auth and SSO, permissions, data governance, observability and reliability on an organizational scale. * Working knowledge of the data layer behind good agents - retrieval, vector databases and data engineering pipelines. ## Description Live Our Purpose: You align your values, actions, and goals with Pearson's mission. You inspire others by articulating a compelling vision, connecting daily work to our broader purpose, and making decisions that reflect a deep understanding of our strategy. Simplify the Complexity: You navigate ambiguity and distill complexity into clear, actionable plans. You help teams understand how their work contributes to strategic goals, enabling focus, clarity, and effective execution in a dynamic environment. Carry Our Culture: You foster a culture of customer centricity, high performance, and exceptional collaboration. You lead with integrity, empower others, and build inclusive, high-performing teams that reflect Pearson's values and behaviors. Deliver Results: You set ambitious goals and execute with quality, precision, and speed. You hold yourself and others accountable for outcomes, ensuring alignment with strategic priorities and delivering measurable impact for learners and stakeholders. What You'll Do * Deploy Forward with the Business * Embed directly with teams across Pearson as a forward-deployed engineer - sit with the people doing the work, learn their domain, and build agents into their real workflows. * Move from first conversation to a working system in the field in days, then iterate with users in the loop until the agent earns a permanent place in their day. * Own what you build end to end - discovery, architecture, interface, deployment, and the outcome once it is live. * Build Production-Grade Agentic Applications * Design and ship multi-agent systems that use tools via MCP, coordinate with one another over agent-to-agent protocols, and hand work back to people at the right moment. * Build the interfaces that make agents genuinely usable: generative and streaming UI, human-in-the-loop review, voice and speech-to-speech interaction, and multimodal input. * Work fluently across the agentic framework landscape - LangChain and LangGraph, Deep Agents, CrewAI, Google's Agent Development Kit, the Microsoft Agent Framework, and the OpenAI, Claude and Vercel AI SDKs - and choose the right tool for each problem. * Engineer for Trust * Instrument traces, build eval suites and guardrails, and catch regressions before users do. * Ship into enterprise reality: SSO, permissions, data governance, observability and reliability. * Hold a high bar on craft - clean architecture, strong testing, and systems that hold up under real-world use. * Bring the Frontier In * Evaluate emerging frameworks, models and protocols as they land, and bring into production what is ready for it. * Work agent-first in your own engineering, using tools such as Claude Code, Cursor, Codex and Antigravity with strong review and testing discipline. * Turn what you learn in the field into reusable patterns, components and platform capabilities the rest of AgentOps can build on. Who You Are: * An owner and an entrepreneur - as comfortable defining the problem as solving it. * A forward-deployed builder, energised by sitting with users, learning their domain, and shipping into their real workflows. * A full-stack engineer who treats the interface as part of the product. * An early adopter with shipped work behind you - you form your own view on new frameworks by building with them. * A rigorous engineer who reasons from traces, evals and evidence. * A collaborator who turns ideas into impact, fast. * A believer in the mission - you want the agents you build to free people to do their best, most human work.