You can deploy this model and start earning money today!

roy477/
email-triage-ai

Name: email-triage-ai
Availability: InStock

Views

No views yet

📧 Email Triage OpenEnv

An OpenEnv-compatible environment where an AI agent triages emails by deciding what to do with each one.

Built for the OpenEnv Hackathon – Round 1.

🌍 Environment Description

In a real workplace, people spend hours every day managing email. This environment simulates that task: the agent receives emails one at a time and must choose the correct action for each.

Why email triage?

It's a task every professional does
Actions have clear right/wrong answers
Replies can be evaluated for quality
Scales easily from simple to complex scenarios

🎮 Action Space

Action	Description
`reply`	Send a reply (must include `reply_text`)
`archive`	Save for reference, no action needed
`escalate`	Forward to management immediately
`delete`	Permanently delete the email
`spam`	Mark as spam

json

{"action_type": "reply", "reply_text": "Dear sender, thank you for reaching out..."}
{"action_type": "archive"}
{"action_type": "escalate"}

👁️ Observation Space

Each step returns an observation with these fields:

Field	Type	Description
`email_id`	string	Unique ID for this email
`subject`	string	Email subject line
`sender`	string	Sender email address
`body`	string	Email body text
`success`	bool	Whether the last action was valid
`done`	bool	Whether the episode is finished
`reward`	float	Score for the last action (0.0–1.0)
`message`	string	Human-readable feedback

📋 Tasks

Task ID	Difficulty	Emails	Description
`easy`	⭐ Easy	3	Clearly labelled spam and routine emails
`medium`	⭐⭐ Medium	4	Mix of replies, escalations, and archives
`hard`	⭐⭐⭐ Hard	5	Ambiguous emails requiring nuanced judgment

🏆 Reward Function

1.0 — Correct action taken
0.6–1.0 — Correct reply with keyword quality scoring
0.5 — Correct action type (reply) but no reply text provided
0.3 — Partially correct (e.g. escalated instead of replied)
0.0 — Wrong action
-0.1 — Invalid action type

🚀 Setup & Usage

Run locally

bash

# Install dependencies
pip install -r requirements.txt

# Start the server
uvicorn app:app --host 0.0.0.0 --port 7860

Run with Docker

bash

docker build -t email-triage-env .
docker run -p 7860:7860 email-triage-env

API Endpoints

Method	Endpoint	Description
GET	`/`	Health check
POST	`/reset?task_id=easy`	Start a new episode
POST	`/step`	Take one action
GET	`/state`	Get current state

Example usage

python

import requests

BASE = "http://localhost:7860"

# Start easy task
obs = requests.post(f"{BASE}/reset", params={"task_id": "easy"}).json()
print(obs["subject"])  # "You won a prize!!!"

# Take action
result = requests.post(f"{BASE}/step", json={"action_type": "spam"}).json()
print(result["reward"])  # 1.0

🤖 Running the Baseline Inference

bash

export API_BASE_URL="https://api.openai.com/v1"
export MODEL_NAME="gpt-4o-mini"
export HF_TOKEN="your-api-key"
export ENV_URL="http://localhost:7860"

python inference.py

Baseline Scores (gpt-4o-mini)

Task	Score
easy	~0.95
medium	~0.80
hard	~0.70
overall	~0.82

📁 Project Structure

email_triage_env/
├── app.py           # FastAPI server
├── environment.py   # Core environment logic
├── models.py        # Pydantic data models
├── tasks.py         # Task definitions and graders
├── inference.py     # Baseline inference script
├── openenv.yaml     # OpenEnv metadata
├── requirements.txt
├── Dockerfile
└── README.md

👥 Team

VisionX — OpenEnv Hackathon 2025

roy477/email-triage-ai

📧 Email Triage OpenEnv

🌍 Environment Description

🎮 Action Space

👁️ Observation Space

📋 Tasks

🏆 Reward Function

🚀 Setup & Usage

Run locally

Run with Docker

API Endpoints

Example usage

🤖 Running the Baseline Inference

Baseline Scores (gpt-4o-mini)

📁 Project Structure

👥 Team

roy477/
email-triage-ai