Ethan Park

Designs prompt wrappers that enforce strict JSON schema compliance with validation rules and error recovery.

Creates meta-prompts that generate task-specific prompts with built-in quality checks and self-improvement mechanisms.

Aggregates multiple LLM responses using self-consistency voting to identify majority answers with confidence scoring.

Creates detailed expert persona system prompts with authentic professional characteristics and calibrated confidence levels.

Evaluates and ranks multiple solution branches in tree-of-thought reasoning, recommending pruning and exploration strategies.

Transforms standard prompts into chain-of-thought versions with explicit reasoning scaffolds and logical sequencing.

Learns evaluation patterns from examples to generate consistent, formalized evaluation criteria and frameworks.

Analyzes prompts for clarity issues and ambiguity without examples, providing actionable improvement recommendations.