Stochastic Modeling Intern

About the Role

This is a paid internship for a graduate student or advanced undergraduate. You will work directly with Canyon Code's CEO on the data side of CanyonOS, characterizing how multi-agent workflows actually run and building statistical models for prediction.

Key Responsibilities

• Design and run experiments over agentic workflows: grids of tasks, prompts, and configurations with repeated runs, on public benchmarks and on CanyonOS traces.
• Characterize run-to-run variability in tokens, latency, and tool calls, and how it depends on inputs, prompts, and configuration.
• Build probabilistic models of workflow behaviour, such as state-transition models and Bayesian hierarchical models estimated from small samples, and test their predictions on held-out runs.
• Write static analysis over Python agent code to extract its structure: agents, calls between them, loops, retries, and prompt templates.
• Reproduce and extend results from recent papers on agent token consumption, using public trajectory datasets.
• Work with the engineering team to bring what works into CanyonOS's telemetry and dashboard.

Requirements & Tech Stack

• Graduate student or advanced undergraduate in statistics, applied mathematics, computer science, electrical engineering, or operations research.
• Solid probability and statistics: Markov chains, Bayesian estimation, hierarchical models, credible intervals, the bootstrap. Ability to derive results is essential.
• Experience designing and analyzing experiments: factorial grids, repeated runs, effects on a log scale, held-out evaluation.
• Fluent in Python with numpy, scipy, and pandas.
• Comfortable working with large trace datasets.
• Experience with a probabilistic programming tool such as PyMC, Stan, or NumPyro.

Plus

• Experience building or closely using LLM agents (in a framework or plain Python) and understanding of prompts, tool calls, context growth, and token accounting.
• Ability to run agent benchmarks at scale: SWE-bench or similar harnesses, Docker, batch execution, cost tracking.
• Ability to read other people's Python code and reason about its control flow; experience with Python's ast module is a plus.
• Familiarity with queueing theory or performance modelling.
• Familiarity with LLM serving stacks such as vLLM or SGLang, and understanding of prompt and KV caching implications for cost.