Core Agent Engineer

Adobe•San Francisco, CA

About The Position

Adobe is hiring a hands-on full-stack engineer to help build an AI-native Agentic Product that brings generative models, creative intelligence, and Adobe capabilities together for creators. The role involves developing the agent across its full stack and Adobe’s target platforms, including desktop, mobile, web, and embedded surfaces. The engineer will help establish the team’s technical foundations and practices, taking capabilities from prototypes to robust, scalable, maintainable production systems. The work spans LLM integration and agentic reasoning, APIs, client runtimes, platform delivery, model orchestration, tool use, multi-step task execution, context and memory management, and evaluation infrastructure. The engineer will also advance AI-assisted engineering practices, developer experience, and systematic quality measurement for agent behavior. In partnership with Product, Engineering, platform, infrastructure, data, and security teams, this person will help deliver a reliable product that meets Adobe’s quality, trust, data protection, observability, and operational requirements.

Requirements

Hands-on experience with large language models, including LLM APIs, prompt engineering, context window management, retrieval-augmented generation, and operating LLM-backed systems in production.
Experience building and shipping agentic systems, including tool use, multi-step reasoning, orchestration frameworks, memory and context persistence, and understanding of failure modes.
Hands-on experience using AI-assisted engineering tools such as Claude Code, Codex, Cursor, or similar in real development workflows.
Full-stack engineering depth to contribute across backend services, client SDKs, and platform-specific integrations.
Strong engineering craft, including writing readable, well-structured, production-quality code and establishing effective testing and local development practices.
Clear communication skills and ability to discuss technical tradeoffs with product leads and senior technical stakeholders, flag issues early, and collaborate across disciplines.
Ability to work in a fast-moving, ambiguous environment where models, frameworks, and user expectations evolve rapidly.
Experience collaborating cross-functionally with product, design, platform, and security teams to ship user-facing AI products.
Familiarity with quality, trust, and reliability requirements for AI features serving large consumer and enterprise audiences, including responsible AI considerations for generative and agentic systems.

Responsibilities

Own full-stack implementation of the Agentic Product, including LLM integration, agentic reasoning layers, API surfaces, client runtimes, and delivery across desktop, mobile, and web.
Help turn prototype-quality capabilities into robust, scalable, maintainable production systems.
Implement high-leverage components such as model orchestration, tool use, multi-step task execution, context and memory management, and agent evaluation infrastructure.
Promote AI-assisted engineering practices using tools such as Claude Code, Codex, and emerging agentic development platforms.
Develop workflows in which AI is a first-class collaborator in engineering.
Improve developer experience through tooling, local development loops, CI/CD ergonomics, and team scaffolding.
Evaluate evolving LLM capabilities and frameworks and recommend when and how to adopt them.
Build and maintain agent evaluation and quality frameworks, including evals, benchmarks, and feedback loops to measure progress and catch regressions.
Partner with Engineering and Product leadership to translate product goals into technically grounded delivery plans, raise constraints early, and propose implementation options.
Collaborate with platform, infrastructure, data, and security teams on guardrails, data protection, observability, and operational requirements.
Set a strong example through excellent code, high-signal code reviews, and sharing technical knowledge in agentic engineering.