Agent Engineer San Francisco CA

AHU Technologies; Washington (District of Columbia); On-site; USD 180,000 - 240,000; Full time

Job summary

AHU Technologies in San Francisco, CA is seeking an experienced Agent Engineer to build and scale the agent and workflow layer behind a cloud-based creative platform. The role requires hands-on experience shipping production AI or backend systems and designing reliable model orchestration.

You will develop intelligent workflow systems that integrate AI models, tools, and services across image, video, audio, text, and other creative workflows, ensuring scalable, reliable operations.

Qualifications

• 3+ years shipping AI, agent-based, backend, or production software systems.
• Strong Python and/or TypeScript skills.
• Experience with backend development, databases, queues, and production reliability practices.
• Hands-on experience with agent architectures (planning, tool use, memory, retrieval, context management, workflow execution).
• Experience orchestrating multiple AI models and providers (routing, retries, fallbacks, latency, cost management).
• Experience with evaluation frameworks and A/B testing environments.
• Ability to work on-site in San Francisco.

Responsibilities

• Build and maintain agent runtimes supporting planning, tool usage, memory, retrieval, context management, and workflow execution.
• Orchestrate multiple AI models and providers across generation workflows.
• Design routing, retry mechanisms, fallbacks, latency optimization, cost controls, and failure recovery systems.
• Develop workflow memory systems preserving project context, assets, templates, style preferences, and previous outputs.
• Implement evaluation frameworks, regression testing, benchmarking, human review workflows, and quality metrics.
• Improve production reliability through queue management, monitoring, cancellation handling, progress tracking, and graceful recovery mechanisms.
• Collaborate with cross-functional teams to deliver scalable and user-friendly AI workflow solutions.

Skills

Python

TypeScript

Backend development

Observability

Production reliability

Agent architectures

AI workflow orchestration

Tools

Databases

Queue systems

Observability tooling

Job description

Position: Agent Engineer

Visa Status: US Citizens and Green Card Holders (Only)

Location: San Francisco, CA (On-site)

Employment Type: Full-Time

Experience Required: 10+ Years

About the Role

We are seeking an experienced Agent Engineer to build and scale the agent and workflow layer behind a cloud-based creative platform. The ideal candidate has experience shipping production AI or backend systems, understands agent runtime behavior, and can design reliable model orchestration beyond basic prompt-based applications.

This role focuses on developing intelligent workflow systems that integrate multiple AI models, tools, and services across image, video, audio, text, and other creative workflows.

Key Responsibilities

• Build and maintain agent runtimes supporting planning, tool usage, memory, retrieval, context management, and workflow execution.
• Orchestrate multiple AI models and providers across various generation workflows.
• Design routing, retry mechanisms, fallbacks, latency optimization, cost controls, and failure recovery systems.
• Develop workflow memory systems that preserve project context, assets, templates, style preferences, and previous outputs.
• Implement evaluation frameworks, regression testing, benchmarking, human review workflows, and quality metrics.
• Improve production reliability through queue management, monitoring, cancellation handling, progress tracking, and graceful recovery mechanisms.
• Collaborate with cross-functional teams to deliver scalable and user-friendly AI workflow solutions.

Required Qualifications

• 3+ years of experience building and shipping AI, agent-based, backend, infrastructure, or production software systems.
• Strong proficiency in Python and/or TypeScript.
• Experience with backend development, databases, queues, observability, and production reliability practices.
• Hands-on experience with agent architectures, including planning, tool use, memory, retrieval, context management, and workflow execution.
• Experience orchestrating multiple AI models and providers, including routing, retries, fallbacks, latency optimization, and cost management.
• Experience building evaluation frameworks, regression testing systems, quality metrics, or A/B testing environments.
• Strong problem-solving skills and ability to work in a fast-paced environment.
• Ability to work on-site in San Francisco.

Preferred Qualifications

Creative AI Systems

• Experience working with image, video, audio, TTS/STT, lipsync, upscaling, or related generative AI pipelines.
• Experience maintaining reliable generation workflows at scale.

Workflow & Runtime Systems

• Experience with workflow engines, node-based systems, visual programming environments, automation runtimes, or similar platforms.
• Strong understanding of queues, state management, retries, and recovery mechanisms.

Quality & Evaluation

• Experience designing AI evaluation systems, regression testing, human review workflows, and benchmarking frameworks.
• Ability to measure and improve AI output quality in production environments.

What Success Looks Like

• Building reliable agent workflows that scale efficiently in production.
• Improving workflow quality, reliability, and user experience.
• Delivering robust orchestration across multiple AI systems and providers.
• Establishing strong evaluation and monitoring practices for AI behavior.