Kevin Chen

- Evaluate LLM outputs for whether responses are correctly grounded in source documents — checking factual accuracy, correctness, and failure-mode handling on document-grounded Q&A tasks, achieving 3.9/5 quality rating.
- Serve as a reviewer on the project, auditing peer evaluations for accuracy and adherence to grading and formatting guidelines before final submission.
- Evaluated agentic AI-powe

- Trained a flagship LLM-based CLI through prompt engineering, comparative model evaluation, and model ranking — assessing whether AI-generated code met production-readiness and correctness standards before acceptance.

- Engineered prompts and rubrics to evaluate, rate, and compare multimodal AI responses across text, image, and video tasks.
- Trained Grok's video understanding capabilities by breaking down video clips into cinematic components.
- Promoted to reviewer across 3 projects — reviewing peers' work, providing detailed feedback, and correcting errors against project guidelines.
- Achieved a 6.6/7 overa

- Debugged AI-generated code and evaluated, analyzed, and compared AI model responses across multiple programming languages.
- Trained AI SWE agents to extract requirements and analyze GitHub pull requests for quality evaluation.
- Promoted to reviewer across multiple projects — evaluating tasks and giving detailed feedback to peers.
- Engineered prompts in Simplified Chinese, ensuring proper loca

- Led end-to-end Android development as sole developer on an AI chat application, independently owning UI, and feature implementation.
- Collaborated with cofounders to identify and develop features.

- Transitioned from customer service to Programming.
- Self-taught Android and web, and built apps.
- Also enrolled in full-stack web coding bootcamp,

- Took customer orders, packaged food and beverage, and maintained sanitary conditions of establishment.

Completed a Bachelor of Arts in General Linguistics at Queens College.