Eliza

ELIZA is a classic rule-based conversational agent originally developed at MIT in the 1960s, providing a pattern-matching Rogerian therapist baseline.

Overview

Property	Value
Key	eliza
Type	Rule-based
Focus	Rogerian Therapy

Key Features

• Pattern matching: Uses regex patterns to generate contextual responses
• Reflection: Reflects client statements back as questions
• Open-ended questions: Encourages elaboration without leading
• No LLM required: Fully deterministic, no API key needed

How It Works

ELIZA implements the classic DOCTOR script. On each turn it scans the input for matching patterns (e.g. "I feel X" → "Why do you feel X?"), applies a reflection transform, and falls back to generic prompts when no pattern matches.

Usage

CLI

uv run python -m examples.simulate therapist=eliza

Python

from patienthub.therapists import get_therapist

therapist = get_therapist(agent_name="eliza", lang='en')

response = therapist.generate_response("I feel very anxious.")
print(response)

Configuration

ELIZA uses a fixed rule-based system and does not require any configuration parameters.

Use Cases

• Baseline comparisons with LLM-based therapists
• Historical/educational demonstrations of early conversational AI
• Low-resource environments where LLM access is unavailable
• Research on rule-based vs. LLM therapeutic interaction