Ethan Park

Generates improvement hypotheses from performance analysis with prioritization and validation plans.

Designs clean interfaces between decomposed subtasks with proper schemas and data flow specifications.

Creates multi-expert panel discussions with defined personas, interaction rules, and consensus mechanisms.

Verifies each step in chain-of-thought reasoning for logical validity and proper prerequisites.

Scores semantic similarity between candidate texts and reference examples with feature-level analysis.

Builds zero-shot sentiment analysis prompts with configurable scales and nuanced mixed sentiment handling.

Adapts prompts for different audiences by adjusting vocabulary, examples, and complexity levels.

Implements prompt injection prevention with input sanitization and detection mechanisms.

Analyzes response diversity in self-consistency sampling to optimize temperature and sampling parameters.

Synthesizes final solutions from multiple tree-of-thought branches by combining best elements.

Validates that generated content authentically represents the specified persona across multiple dimensions.

Generates structured counterarguments to test reasoning robustness and identify weak premises.