Software Engineer: Self Improvement

Big Wombat LLC; United States; Remote; USD 215,000 - 325,000; Full time

Benefits offered by this job

Equity

Remote across US/EU time zones

Competitive salary

Job summary

Big Wombat LLC in the United States is seeking a software engineer to build product systems that help an AI assistant learn from everyday work, with emphasis on making improvements usable for non-technical users.

You will design end-to-end pipelines to gather evidence, derive skills and memories, evaluate impact, and provide clear controls over what is retained and applied by the assistant.

The role is full-time and remote across US/EU time zones, with base salary and equity offered.

Qualifications

• Software engineering experience spanning product systems and backend or agent workflows.
• Ability to build and evaluate systems that learn from usage, corrections, or task outcomes.
• Product judgment for making complex AI behavior understandable and controllable to non-technical users.
• Careful handling of relevance, limited context, and privacy when retrieving or sharing user-specific information.

Responsibilities

• Create an end-to-end process for gathering evidence from work, proposing improvements, evaluating them, and applying those that help.
• Build the runtime for finding, retrieving, loading, and combining relevant skills and memories for a task.
• Develop pipelines that derive new or updated skills and memories from session history.
• Design review experiences that let people accept, edit, or reject proposed learnings.
• Evaluate improvements against real tasks, detect when they become stale or harmful, and support correcting or retiring them.
• Explore ways to share useful learnings across a team while protecting private context.

Skills

Software engineering

Backend systems

Agent workflows

Product systems

Job description

Role overview

Build the product systems that help an AI assistant improve from people’s everyday work. The role covers the full learning cycle: detecting useful patterns and corrections, turning them into skills or memories, evaluating whether they improve results, and giving people clear control over what the assistant retains and applies. The experience is intended for non-technical users, so improvements should work without requiring prompt writing or configuration.

Responsibilities

• Create an end-to-end process for gathering evidence from work, proposing improvements, evaluating them, and applying those that help.
• Build the runtime for finding, retrieving, loading, and combining relevant skills and memories for a task.
• Develop pipelines that derive new or updated skills and memories from session history.
• Design review experiences that let people accept, edit, or reject proposed learnings.
• Evaluate improvements against real tasks, detect when they become stale or harmful, and support correcting or retiring them.
• Explore ways to share useful learnings across a team while protecting private context.

Requirements

• Software engineering experience spanning product systems and backend or agent workflows.
• Ability to build and evaluate systems that learn from usage, corrections, or task outcomes.
• Product judgment for making complex AI behavior understandable and controllable to non-technical users.
• Careful handling of relevance, limited context, and privacy when retrieving or sharing user-specific information.

Benefits and work setup

• Full-time engineering role, remote across US or EU time zones.
• US-based hires are offered a stated base salary range of $215,000–$325,000 and equity.