Meta: An AI Researcher at $250,000 Shares Her Journey

Ruiyu Li, 26 years old and based in Menlo Park, claims to work as an AI researcher at Meta and reports a compensation exceeding $250,000, including salary, bonuses, and stock options. Her journey includes eight academic labs, a startup with fewer than ten employees, and an internship at Microsoft, culminating in a video generation pipeline in production. She asserts that verifiable publications and code matter more than lists of experiences. Her current projects focus, she says, on large-scale recommendation, AI agents, and post-LLM optimization.

Evidence in Production: A Video Pipeline and a Refined Specialty

During her internship at Microsoft, Ruiyu Li claims to have contributed to generative research and Bing Copilot, delivering a complete video generation pipeline based on a multimodal LLM, which was put into production in Bing Search. At the startup, she says she worked on automating video editing using AI innovations that were considered state-of-the-art at the time. She believes that these two internships allowed her to refine a specialty around large-scale AI systems in production and video generation, and that the skills acquired proved directly useful in a full-time position.

What Li Looks for in Candidates: Consultable Work

When reviewing applications, Ruiyu Li says she checks for the existence of concrete research and projects. She believes that a simple list of experiences is not sufficient in a competitive market and finds published papers or open-source code that employers can review to be more convincing. She encourages accumulating as much practical experience as possible, particularly recommending an internship at a startup or publishing research. She reports having led the production of AI innovations during an internship at a startup with fewer than ten people, working directly with the CEO and CTO, and then at Microsoft. According to her, increasing the number of projects accelerates learning and adaptation to trends.

Education: CMU in 2023, Eight Labs, and a Focus on Applied Research

Ruiyu Li indicates that she graduated in 2022 and then completed a master's degree in machine learning at Carnegie Mellon University in 2023, with numerous advanced projects allowing her to participate in cutting-edge research. She reports having conducted work in eight AI labs covering various fields, an experience that, she says, helped her publish. Coming from an undergraduate background in computer science, she states that she wanted to deepen her knowledge and focus on applied research in AI. For those specifically targeting AI, she recommends graduate school.

Recruitment Under Pressure: Three Offers, Microsoft, Then Meta

Ruiyu Li describes the job search period as challenging for tech and believes that job fairs and conferences did not help her. Nevertheless, she claims to have received three offers from major companies, first joining Microsoft and then arriving at Meta. She draws practical lessons from this journey about what helps secure offers and considers her internships to be crucial. She also asserts that many employers expect proof of technical experience and industry-level projects.

At Meta: Large-Scale Recommendation, AI Agents, and Post-LLM Optimization

Ruiyu Li indicates that she holds a position as an artificial intelligence researcher at Meta, with a primary focus on AI and machine learning. She specifies that she works on training models for large-scale recommendation systems, creating native AI agents to improve efficiency, as well as optimizing training and inference after LLM. She mentions having acquired expertise in video generation and large-scale AI systems, and clarifies that she implements these skills on production systems requiring a balance of model quality, scalability, reliability, efficiency, and quantifiable impact. Based in Menlo Park, 26 years old, she reports a compensation exceeding $250,000, composed of base salary, bonuses, and stock options. She adds that, despite the assistance of AI, mastery of algorithms and code remains essential in her view.