About the role
The Opportunity Adobe empowers individuals and organizations to create exceptional content effortlessly. The AI for Engineering team builds a scalable, production-grade AI platform that powers creativity across design, imaging, motion, and personalization. We are seeking a Staff Engineer - AI for Engineering to design and build the next-generation AI systems that bring intelligent, adaptive, and agentic experiences to Adobe Express. This role is ideal for an engineer who deeply understands how modern LLMs and generative models behave in real-world systems — including reasoning patterns, tool use, prompt dynamics, failure modes, and orchestration strategies — and can translate that understanding into scalable, production-ready platforms. You will define and build the end-to-end architecture that enables Agentic AI solutions solving real engineering problems— spanning model orchestration, tool integration, memory systems, inference services, data flows, evaluation loops, and real-time decision systems. This is a systems-first AI role focused on building intelligent platforms. What You’ll Do A core part of this role is building scalable Agentic AI platforms that empower developers, significantly boost engineering productivity, and accelerate AI adoption across the organization. Architect and evolve the AI platform powering Adobe Engineering — with a strong emphasis on Agentic AI systems and LLM-native architectures. Design and implement scalable orchestration layers that coordinate LLMs, tools, APIs, memory stores, and multi-step reasoning workflows. Build production-grade agent frameworks that support planning, task decomposition, tool invocation, multi-agent collaboration, and persistent memory. Develop high-performance inference and runtime systems with strong guarantees around latency, reliability, observability, and cost efficiency. Design evaluation and feedback systems that measure reasoning quality, task success, hallucination rates, and agent behavior — enabling rapid iteration and continuous improvement. Integrate first-party and third-party foundation models into cohesive, adaptive systems using routing, model selection, guardrails, and fallback strategies. Design data flows, session-level intelligence, and contextual memory systems that allow agents to operate coherently across interactions. Partner with applied research, product, and platform teams to bring intelligent agentic capabilities into real developer-facing experiences. Drive architectural strategy for AI for Engineering — connecting models, reasoning engines, tools, and data streams into adaptive AI systems. Mentor senior engineers in modern AI system design, LLM orchestration patterns, and agent platform architecture. What You’ll Bring 10+ years of experience building large-scale distributed systems, AI platforms, or intelligent service architectures. Deep understanding of how LLMs behave in production environments — including prompting strategies, reasoning chains, tool usage, grounding techniques, hallucination mitigation, guardrails, and evaluation patterns. Strong experience building AI-powered systems using LLM orchestration frameworks, model routing strategies, and multi-model pipelines. Hands-on experience designing agentic systems — including reasoning loops, memory persistence, tool integration, state management, and multi-agent coordination. Proven expertise in building scalable, cloud-native, microservices-based architectures with strong observability and reliability. Experience designing evaluation systems for generative AI quality, task completion, and behavioral robustness. Proficiency in TypeScript, Python and at least one systems language (Java, Go, C++), with experience building production AI services. Strong systems thinking — ability to connect model capabilities, runtime constraints, and product
