Synthetic Data Engineer

TypeSafe AI•San Francisco, CA

3d•Onsite

About The Position

TypeSafe is a frontier model lab focused on building reliable and general AI systems for economically valuable automation. The company's mission is to usher in a new era of Transformative Artificial Intelligence (TAI). They are rethinking the LLM stack to build a new kind of general frontier model designed for real-world reliability, decision-making, and autonomy in production. The synthetic data engineers will support the core training stack of their models with data science, creativity, and engineering skills, contributing to the 'secret sauce' of TypeSafe by figuring out how models can provide production-ready reliability.

Requirements

5+ years of professional software engineering, or equivalent professional expertise.
Strong coding ability and product intuition.
Enjoy thinking beyond the code to how systems are used in the real world.
Maintain high attention to detail and a strong bar for quality.
Can describe complex problems succinctly and clearly.
Care deeply about our mission and are driven to translate it into real-world impact.
Thrive in ambiguous problem spaces and take satisfaction in finding creative solutions.
Have hands-on experience implementing LLMs and understand their capabilities and limitations.

Responsibilities

Design and build reliable, production-grade AI systems that integrate large language models and proprietary models into real-world workflows.
Create high-leverage datasets, products, and user interfaces that unlock new capabilities and use cases on top of the model.
Run rigorous analyses and experiments to understand how data, training choices, and targeted interventions impact model behavior and performance.
Develop and own evaluation frameworks to measure quality, reliability, and emergent characteristics across model iterations.
Partner closely with product managers and customers to translate real-world needs into concrete model and system improvements.
Continuously iterate, learn, and co-discover new techniques for building exceptional, trustworthy AI products.