Principal AI Taxonomy Engineer

Job Summary

Red Hat's Global Engineering Team is looking for an AI taxonomy engineer to join the newly-formed InstructLab and Granite team. This role offers an exciting opportunity to contribute to the foundation of our AI projects by ensuring well-structured and meaningful data organization. A successful candidate will play a crucial role in enhancing the efficiency and effectiveness of our AI models through robust taxonomy management.

As a taxonomy engineer, you will be responsible for developing, maintaining, and optimizing classification systems that support our AI models and data organization efforts. In this role, you will be responsible for reviewing and categorizing incoming data, issues, or tasks and prioritizing these items based on urgency, importance, and relevance. Effective taxonomy engineers will collaborate closely with data scientists, machine learning engineers, and domain experts to align taxonomies with project requirements, and to understand the effectiveness of taxonomies in supporting AI model performance.

This is a fast moving area, so the ideal candidate will have a collaborative mindset and work effectively in cross-functional teams in an open organization where contributions are valued and expected from all levels. As this is a fast-moving area of opportunity for Red Hat, the ability to communicate productively and effectively with team members, stakeholders, and Red Hat leadership is critical. Strong organizational skills are imperative for success in this role.

This position reports directly to the Manager of Software Engineering for InstructLab. This position must have the ability to partner collaboratively in our Boston, MA office at least 3 days per week. 

Primary Job Responsibilities (what you’ll do)

• Evaluate and optimize triaging processes and documentation over time, ensuring the team’s processes are efficient and up-to-date.

• Triage pull requests (PRs) and route tasks to appropriate team members or departments; identify trends in incoming data.

• Develop and code software programs, algorithms, and automated processes to cleanse, integrate, and evaluate large datasets from multiple disparate sources.

• Ensure consistency and accuracy in data labeling and categorization, continuously updating and refining taxonomies as new categories or concepts emerge.

• Help maintain data quality by filtering out low-quality or unsuitable inputs; as needed, escalate actions in response to sensitive or inappropriate content.

• Create and maintain documentation for taxonomy changes, usage guidelines, and best practices; monitor repository cleanliness and take action to maintain order (e.g., close dated PRs and irrelevant issues). 

• Maintain documentation regarding triage and taxonomy status, as well as skills and knowledge contribution process enhancements.

• …