AI Model Benchmarking Intern (Iowa-Based)

What You'll Do6

• Survey new LLM and VLM releases and judge which ones fit SureHost product use cases.
• Build evaluation datasets, test cases, and scoring criteria that someone else can rerun without guessing.
• Score models on accuracy, reliability, latency, and cost, along with other factors the product depends on.
• Trace failure patterns, including false positives and false negatives on the issues operators flag.
• Present findings and model recommendations to the AI Chief Architect and the product side.
• Keep benchmark results and documentation current as APIs, models, and requirements shift.

Requirements7

• Coursework or personal projects involving LLMs, VLMs, generative AI tools, or model APIs.
• Evidence-based reasoning and comfort working with structured data and evaluation criteria.
• Working Python or data-analysis skills.
• A habit of documenting methods and results so another person can follow them.
• Patience for repeatable experiments and edge cases, plus the judgment to tell a strong demo from dependable performance.
• Clear communication, early questions, and careful handling of customer, property, and company information.
• Based in Iowa.

Who Should Apply

A student or early-career builder based in Iowa who has spent time with LLMs, VLMs, or model APIs and wants to test them against real product footage. You need enough Python and data sense to build repeatable evaluations, plus the patience to chase edge cases until the numbers mean something.

Salary Insight