SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Salary: USD 250,000 - 450,000 / annual
Abundant is a research lab focused on identifying approaches to solve long-horizon tasks. As a Member of Technical Staff with a focus on Safety, you will work directly with the founding team to ensure research is conducted responsibly and benefits society.
You will embed AI safety principles into all company workflows and products. Your responsibilities include:
- Foundational Capability Research & Execution: Architect and execute research nested within Abundant's focus areas, particularly long-horizon tasks and simulation environments.
- Drive Industry Leadership: Establish Abundant as a leader in AI post-training safety, shaping the conversation around responsible AI development.
- Build Novel Tooling: Identify and develop tooling that advances safety goals and enables better evaluation and monitoring of AI systems.
You will have the opportunity to work on high-stakes problems at the intersection of AI research and safety, including advanced post-training methodologies, distillation techniques, and evaluation frameworks. The role involves both foundational research and direct productization of safety-focused innovations.
REQUIREMENTS:
- PhD in a related field (e.g., Computer Science, Machine Learning, NLP) with a world-class publication record (NeurIPS, ICML, ICLR) preferred.
- Proven experience shipping research directly to production and live systems, specifically in advanced post-training, distillation, and high-stakes evaluation methodologies.
- Expertise in large-scale agent systems, including orchestration frameworks, tool APIs, distributed execution, observability, and logging infrastructure.
- Demonstrated velocity to master complex fields and proven track record in research productization.
- Thoughtful, strategic perspective on societal and safety impacts of deploying general-purpose AI systems.
- Experience advising on or shaping governmental policy related to AI safety and governance is preferred.
