verahession 0.1.9

Python client library for Vera AI chatbot by Hession Dynamics

Verahession

Verahession is a lightweight Python interface for interacting with V.E.R.A. (Vocally Enhanced Research Assistant), an AI-powered conversational agent. This module provides simple tools to connect to the VERA API, send messages, rewrite input, and run local intent classification using your own training data.

---

---

🔧 Features

• 🧠 Send and receive messages using an API-based LLM
• ✍️ Rewrite user input intelligently
• 🏋️‍♂️ Train a local intent classification model
• 🔍 Detect user intent with confidence scoring

---

📦 Installation

git clone https://github.com/yourusername/verahession.git
cd verahession
pip install -r requirements.txt

---

🔑 API Key

To use the Vera API (LLM), you need an API key.
Visit hessiondynamics.com to request your key.

---

🚀 Example Usage

from verahession.api import vera_interface
from verahession.assistant import *

# Initialize Vera interface
vera = vera_interface(API_KEY="your_api_key_here", AGENT_NAME="Brian", USER_NAME="Jack")

# Send a message
message = input("You: ")
result = vera.send(message)
print("Vera says:", result.get("response", result))

# Rewrite a message (or text, whatever)
message = input("Text: ")
result = vera.rewrite(message)
print("Vera rewrites:", result.get("response", result))

# Train the intent classifier
bot_trainer = trainer("./intents.json", "./model.pth")
bot_trainer.train()

# Classify a new input
classifier = Classifier("./model.pth")
message = input("You: ")
intent, confidence = classifier.classify(message)
print(intent)
print(confidence)

---

📁 File Structure

• verahession/api.py – Vera API interface (LLM)
• verahession/assistant.py – Local trainer & classifier tools
• intents.json – Training data for intent classification
• model.pth – Output model file after training

---

📄 License

MIT License

---