SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Salary: USD 100,000 - 350,000 / annual
Hercules is seeking a Staff Software Engineer to lead the development of its AI coding agent platform. You will own the design, implementation, and optimization of Hercules' AI agent system, including both online and offline evaluation frameworks. This role involves researching and implementing novel approaches to improve agent performance across cost, speed, and accuracy dimensions, leveraging access to 1M+ repositories and millions of chat messages for experimentation and validation.
Key responsibilities include:
- Building and maintaining the core AI coding agent that powers Hercules' platform
- Designing and owning comprehensive evaluation systems (online and offline) to measure agent quality
- Researching and implementing advanced context and harness engineering techniques to optimize agent performance
- Contributing to architectural decisions across the full stack
- Shipping features and improvements directly to customers with a bias toward speed and iteration
The role requires deep expertise in modern AI engineering practices. You should have hands-on experience with cutting-edge context management techniques including summarization, compression, prompt caching, subagents, memory systems, RAG, skills frameworks, and tool orchestration. The field evolves rapidly, and you should have battle-tested opinions on what works at scale.
Hercules operates with a strong in-office culture in San Francisco (Kearny and Bush). The founding team works together daily. Remote arrangements are considered only for exceptional candidates willing to work US time zones and travel to the office regularly.
The tech stack includes TypeScript, React, Postgres, Redis, Temporal, Clickflare, AWS, and integrations with OpenAI, Anthropic, and Google models. You should be comfortable with modern DevOps practices (Docker, IaC with CDK/Terraform), observability (Otel), and working in monorepos.
Hercules values shipping fast, striving for excellence in execution, and operating with hunger to break through obstacles. You should have a track record of independently shipping features to customers and possess strong product/design instincts.
REQUIREMENTS:
- Exceptional software fundamentals: system design, architecture, data modeling, API design
- Proven ability to write clean, maintainable code and work effectively in monorepos
- Strong product and design sense with a track record of independently shipping features
- Deep, hands-on experience with modern context and harness engineering best practices at scale (summarization, compression, prompt caching, subagents, memory, RAG, skills, tools, async/deferred tools, etc.)
- Proficiency in TypeScript, React, and the core Hercules stack (Postgres, Redis, Temporal, containers)
- Ability to leverage AI coding tools to 10x your own workflow
- Comfort with rapid iteration and evolving best practices in AI engineering
