Model Behavior Architect

at Perplexity AI

📍 San Francisco, United States

USD 180,000-260,000 per year

MIDDLE

✅ On-site

Used Tools & Technologies

Not specified

Required Skills & Competences
Tag name is followed by "@" symbol and proficiency level value. About proficiency levels:

1-2 — basic awareness. Minimal hands-on experience, and a rudimentary understanding of the technology's purpose;

3-6 — daily use. Comfortable and regular usage, capable of handling common tasks and challenges related to the technology;

7-9 — you are an expert, you can teach others, you know all the pitfalls and tricks;

10 — exceptional knowledge, comprehensive understanding, and adeptness in all aspects of the technology, including advanced problem-solving. Think twice before claiming or demanding such level.

Python @ 2 Data Science @ 3 Communication @ 3 Debugging @ 2

Details

Perplexity is looking for a Model Behavior Architect to help shape our answer engine. This role collaborates with research, design, and engineering teams to align model behavior with product goals and to evaluate model outputs for consistency and effectiveness.

Responsibilities

Design, refine, and implement context engineering strategies to guide model behavior across use cases.
Identify inconsistencies in model outputs through well-designed research and evaluation projects (internal and production-facing systems).
Develop automated and semi-automated pipelines for evaluating agent effectiveness at scale.
Stay current with the latest alignment and evaluation techniques from both industry and academia.
Collaborate with research, design, and engineering teams to align model behavior with product goals and answer-generation components.

Requirements

Demonstrated passion for AI with specific, related projects you can share.
Familiarity with Python; ability to prototype and debug systems at scale and quickly understand components of answer generation (this is not strictly a production engineering role, but prototyping/debugging skills are required).
High-level understanding of large language models (LLMs), training, and reinforcement learning (RL).
Backgrounds the team has found successful include philosophy, psychology, linguistics, data science, or related fields.
Excellent communication skills, particularly in teaching complex concepts.
Creative problem-solving abilities with a "hacker spirit" for solving puzzles.

Benefits

U.S. full-time employees: comprehensive benefits program including equity, health, dental, vision, retirement, fitness, commuter and dependent care accounts, and more.
International full‑time employees: comprehensive benefits program tailored to region of residence.
Compensation range (U.S.-based positions): $180,000 – $260,000 and the role offers equity. USD salary ranges apply only to U.S.-based positions; international salaries are set based on the local market.