AfterQuery is a frontier AI research lab building training data and evaluation infrastructure used by leading foundation model labs. Backed by Y Combinator, BoxGroup, and top investors from DeepMind, Meta, and Google, the team operates at the cutting edge of model evaluation, reasoning, and agentic systems.
This is a high-impact research internship working directly with AfterQuery’s research team on frontier AI problems. You’ll design experiments, analyze datasets, and help build evaluation frameworks that shape how next-generation AI models are trained and measured.
This role is highly hands-on and fast-paced—ideal for candidates who can think independently, ask strong research questions, and execute quickly.
Design and run experiments focused on model reasoning, agents, and evaluation frameworks
Collect, process, and analyze proprietary datasets for AI model training and benchmarking
Develop and test algorithms related to LLMs, RLHF/RLVR, and agent systems
Support the creation of novel benchmarks and evaluation methodologies
Contribute to research documentation and co-author technical outputs
Present findings internally and collaborate with researchers and other interns
Currently enrolled in an undergraduate or master’s program (CS, AI, Math, or related)
Strong problem-solving ability and curiosity for AI research
Ability to read research papers and extract key insights
Solid programming skills (Python preferred)
Strong communication skills (can explain complex concepts clearly)
Experience with LLMs, reinforcement learning, or agent-based systems
Prior research experience (academic or industry)
Familiarity with NLP or multimodal systems
Experience with data analysis, experimentation, or benchmarking
Publications or contributions to research projects
Flexible (Remote or On-site in San Francisco)
10–40 hours per week, depending on availability
Resume / Project Screening
Take-home research project
Interview with the research team
(Potential) Work trial
Offer
Take the next step in your career journey