Senior Software Engineer (AI Applications)

📋 Description

• • Architect and deliver next-generation autonomous AI agents that power the largest online assessment platform in the United States, directly influencing learning outcomes for over 15 million students each year.
• • Own the end-to-end lifecycle of agentic systems—from whiteboard sketches to production—crafting persistent memory layers, shared state orchestration, and multi-step reasoning pipelines that adapt in real time to every learner.
• • Integrate cutting-edge generative AI capabilities (LLMs, multimodal models, and RAG) into high-throughput web applications that must remain performant and secure while serving peak loads of 1.5 million concurrent users.
• • Establish rigorous evaluation frameworks that measure agent accuracy, safety, and fairness, ensuring that every AI-driven insight delivered to educators meets the highest standards of scientific validity and ethical responsibility.
• • Collaborate with learning scientists, psychometricians, UX researchers, and policy experts to translate complex pedagogical requirements into scalable, production-ready agentic workflows that personalize assessment at the individual student level.
• • Lead technical design reviews and architecture sessions, mentoring senior and mid-level engineers on best practices for secure prompt engineering, tool-calling patterns, and context-window optimization in educational contexts.
• • Champion a culture of experimentation by prototyping novel agent behaviors—such as self-reflection loops, chain-of-thought verification, and human-in-the-loop guardrails—and rapidly iterating based on A/B test results and educator feedback.
• • Drive observability and reliability initiatives, implementing distributed tracing, anomaly detection, and automated rollback strategies that keep AI services resilient during state-mandated testing windows.
• • Contribute to open internal libraries that abstract common agent patterns (tool registries, memory stores, evaluation harnesses) so that multiple product teams can safely and efficiently build AI features on a shared foundation.
• • Influence the strategic roadmap by surfacing emerging research in generative AI, evaluating new model providers, and quantifying cost-performance trade-offs for large-scale inference workloads.
• • Ensure compliance with FERPA, COPPA, and state-level data-privacy regulations by embedding privacy-by-design principles into every layer of the agent stack, from data ingestion to model fine-tuning.
• • Present technical achievements and lessons learned at internal guilds, external EdTech conferences, and peer-reviewed publications, positioning Cambium as a thought leader in responsible AI for education.
• • Foster an inclusive, remote-first engineering culture that values psychological safety, asynchronous communication, and continuous learning, enabling distributed teams to ship high-impact features every sprint.