import gradio as gr import threading import os import pickle import time import json from flask import Flask, request, jsonify from environment import DocumentClassificationEnv from baseline_inference import run_task, load_or_train from agent import TicketAgent from rl_trainer import RLTrainer import numpy as np print("=" * 60) print(" MetaX AI Agent - HACKATHON EDITION") print("=" * 60) print("Loading AI models...") models = {} for d in ["easy", "medium", "hard"]: models[d] = load_or_train(d) print(f" [OK] {d.upper()} model loaded") agent = TicketAgent(model_dir=".") print("[OK] Agent ready") CATEGORIES = { "easy": ["General", "Billing", "Support", "Technical", "HR"], "medium": [ "General", "Billing", "Support", "Technical", "HR", "Legal", "Sales", "Marketing", "Operations", "Complaints", ], "hard": [f"Cat_{i}" for i in range(22)], } api = Flask(__name__) api_envs = {} rl_state = {"trainer": None, "is_training": False} @api.route("/api/reset", methods=["POST"]) def api_reset(): data = request.json or {} difficulty = data.get("difficulty", "easy") seed = data.get("seed", 42) env = DocumentClassificationEnv(task_difficulty=difficulty, seed=seed) obs, _ = env.reset(seed=seed) api_envs["current"] = env return jsonify( { "observation": { k: v.tolist() if hasattr(v, "tolist") else v for k, v in obs.items() } } ) @api.route("/api/step", methods=["POST"]) def api_step(): data = request.json or {} action = data.get("action", 0) env = api_envs.get("current") if env is None: return jsonify({"error": "Call /api/reset first"}), 400 obs, reward, done, _, info = env.step(int(action)) return jsonify( { "observation": { k: v.tolist() if hasattr(v, "tolist") else v for k, v in obs.items() }, "reward": reward, "done": done, "info": info, } ) @api.route("/api/run", methods=["POST"]) def api_run(): data = request.json or {} difficulty = data.get("difficulty", "easy") score, t = run_task(difficulty) return jsonify({"difficulty": difficulty, "score": score, "time": t}) @api.route("/api/agent/process", methods=["POST"]) def api_agent_process(): data = request.json or {} text = data.get("text", "") difficulty = data.get("difficulty", "easy") ticket_id = data.get("ticket_id", None) if not text: return jsonify({"error": "text field required"}), 400 result = agent.process_ticket(text, difficulty=difficulty, ticket_id=ticket_id) return jsonify(result) @api.route("/api/agent/classify", methods=["POST"]) def api_agent_classify(): data = request.json or {} text = data.get("text", "") difficulty = data.get("difficulty", "easy") category, scores = agent.classify(text, difficulty) priority = agent.get_priority(text) return jsonify({"category": category, "priority": priority, "scores": scores}) @api.route("/api/rl/train", methods=["POST"]) def api_rl_train(): global rl_state data = request.json or {} difficulty = data.get("difficulty", "easy") episodes = data.get("episodes", 20) if rl_state.get("is_training"): return jsonify({"status": "Training already in progress"}), 400 env = DocumentClassificationEnv(task_difficulty=difficulty, seed=42) trainer = RLTrainer(env, num_episodes=episodes) rl_state["trainer"] = trainer rl_state["is_training"] = True result = trainer.train_step() eval_result = trainer.evaluate(num_episodes=5) curve = trainer.generate_learning_curve() summary = trainer.get_metrics_summary() rl_state["is_training"] = False return jsonify( { "training_result": result, "evaluation": eval_result, "learning_curve": curve, "summary": summary, } ) @api.route("/api/rl/status", methods=["GET"]) def api_rl_status(): if rl_state.get("trainer") is None: return jsonify({"status": "No training data"}) return jsonify(rl_state["trainer"].get_metrics_summary()) def run_flask(): api.run(host="0.0.0.0", port=7861, debug=False) SAMPLE_TICKETS = { "Billing Complaint": "My invoice shows an incorrect amount! I was charged $500 but should have been charged $75. This is unacceptable!", "Critical Bug": "The application crashes whenever I try to upload a file larger than 10MB. This is critical!", "HR Policy": "I have a question about the company's maternity leave policy. How many weeks of paid leave am I entitled to?", "Login Issue": "I can't log into my account. The password reset link isn't working. Please help!", "Sales Inquiry": "I would like to inquire about upgrading our current subscription to include additional enterprise features.", "Legal Request": "We need legal review of the new vendor contract before signing.", "Frustrated Customer": "I'm extremely disappointed with the service. This is the third time this month I've had issues.", } env_state = {"env": None, "difficulty": "easy", "trainer": None} def create_env(difficulty): e = DocumentClassificationEnv(task_difficulty=difficulty, seed=42) obs, _ = e.reset() env_state["env"] = e env_state["difficulty"] = difficulty model = models[difficulty] cats = CATEGORIES[difficulty] proba = model.predict_proba([obs["content"]])[0] top3 = sorted(enumerate(proba), key=lambda x: -x[1])[:3] explain = "\n".join([f" {cats[i]}: {p * 100:.1f}%" for i, p in top3]) pred_idx = int(model.predict([obs["content"]])[0]) pred_cat = cats[pred_idx] return ( f"Environment Ready | {difficulty.upper()}", obs["content"], f"ML Prediction: {pred_cat}\n\nTop 3:\n{explain}", "", ) def classify_doc(category_idx): e = env_state["env"] if e is None: return "Create environment first!", "", "", "" obs, reward, done, _, info = e.step(int(category_idx)) result = ( f"{'CORRECT' if info.get('is_correct') else 'WRONG'} | Reward: {reward:.2f}" ) if done: acc = info.get("episode_accuracy", 0) result += f"\nEpisode Complete! Accuracy: {acc:.1%}" return result, "Episode complete", "", str(info) difficulty = env_state["difficulty"] model = models[difficulty] cats = CATEGORIES[difficulty] proba = model.predict_proba([obs["content"]])[0] top3 = sorted(enumerate(proba), key=lambda x: -x[1])[:3] explain = "\n".join([f" {cats[i]}: {p * 100:.1f}%" for i, p in top3]) pred_idx = int(model.predict([obs["content"]])[0]) pred_cat = cats[pred_idx] return ( result, obs["content"], f"ML Prediction: {pred_cat}\n\nTop 3:\n{explain}", str(info), ) def process_ticket_ui(ticket_text, difficulty): if not ticket_text.strip(): return "Please enter ticket text!", "", "", "", "", "" result = agent.process_ticket(ticket_text, difficulty=difficulty) cat_out = f"{result['category']}\n Confidence: {result['confidence'] * 100:.1f}%" pri_out = f"{result['priority'].upper()}" dept_out = f"{result['department']}\n {result['email']}\n SLA: {result['sla']}" top3_out = "\n".join([f" {c}: {p * 100:.1f}%" for c, p in result["top3"]]) ref_out = f"{result['ref_id']}\n {result['timestamp'][:19]}" reply_out = result["reply"] return cat_out, pri_out, dept_out, top3_out, ref_out, reply_out def run_rl_training(difficulty, num_episodes): difficulty = difficulty or "easy" num_episodes = num_episodes or 20 try: env = DocumentClassificationEnv(task_difficulty=difficulty, seed=42) trainer = RLTrainer(env, num_episodes=num_episodes) env_state["trainer"] = trainer result = trainer.train_step() curve_img = trainer.generate_learning_curve() summary = trainer.get_metrics_summary() status = f"Training Complete!\n\nResults:\n Episodes: {num_episodes}\n Avg Reward: {result.get('avg_reward', 0):.2f}\n Avg Accuracy: {result.get('avg_accuracy', 0):.1%}" eval_result = trainer.evaluate(num_episodes=5) status += f"\n\nEvaluation Accuracy: {eval_result.get('accuracy', 0):.1%}" return status, curve_img if curve_img else "", str(summary) except Exception as e: return f"Error: {str(e)}", "", "" def load_sample(sample_name): return SAMPLE_TICKETS.get(sample_name, "") def run_evaluation(): rows = [] for d in ["easy", "medium", "hard"]: score, t = run_task(d) rows.append([d.upper(), f"{score:.4f}", f"{score * 100:.1f}%", "Ready"]) return rows def process_image(file_obj): if file_obj is None: return "No image uploaded. Click 'Upload' button, select an image, then click 'Analyze Image'." return f"Image received: {type(file_obj)}. Image analysis requires PyTorch installation." with gr.Blocks(title="MetaX AI - Hackathon Demo") as demo: gr.Markdown("""

MetaX AI Agent

Document Classification | RL Training | AI-Powered

""") with gr.Tab("AI Agent Demo"): gr.Markdown("### AI-Powered Ticket Classification") with gr.Row(): with gr.Column(scale=1): sample_dd = gr.Dropdown( choices=list(SAMPLE_TICKETS.keys()), label="Load Sample" ) ticket_input = gr.Textbox( label="Ticket Text", lines=5, placeholder="Enter customer message...", ) diff_agent = gr.Radio( ["easy", "medium", "hard"], value="medium", label="Difficulty" ) btn_process = gr.Button("Process Ticket", variant="primary", size="lg") with gr.Column(scale=1): cat_out_box = gr.Textbox(label="Category", lines=2) pri_out_box = gr.Textbox(label="Priority", lines=2) dept_out_box = gr.Textbox(label="Routing", lines=2) top3_out_box = gr.Textbox(label="Top Predictions", lines=2) with gr.Row(): ref_out_box = gr.Textbox(label="Reference", lines=1) reply_out_box = gr.Textbox(label="AI Response", lines=8) sample_dd.change(load_sample, inputs=sample_dd, outputs=ticket_input) btn_process.click( process_ticket_ui, inputs=[ticket_input, diff_agent], outputs=[ cat_out_box, pri_out_box, dept_out_box, top3_out_box, ref_out_box, reply_out_box, ], ) with gr.Tab("RL Training"): gr.Markdown("### Train RL Agent") with gr.Row(): with gr.Column(scale=1): diff_rl = gr.Radio( ["easy", "medium", "hard"], value="easy", label="Difficulty" ) episodes_slider = gr.Slider(10, 100, value=30, step=5, label="Episodes") btn_train = gr.Button("Start Training", variant="primary", size="lg") with gr.Column(scale=2): train_status = gr.Textbox(label="Status", lines=8) learning_curve_img = gr.Image(label="Training Progress") btn_train.click( run_rl_training, inputs=[diff_rl, episodes_slider], outputs=[train_status, learning_curve_img, train_status], ) with gr.Tab("Multi-Modal"): gr.Markdown("### Image Upload (Experimental)") gr.Markdown( "Note: Full image analysis requires PyTorch. Install: `pip install torch transformers`" ) with gr.Row(): file_input = gr.File( label="Upload Image", file_count="single", file_types=["image"] ) image_btn = gr.Button("Analyze Image") image_output = gr.Textbox(label="Result", lines=4) image_btn.click(process_image, inputs=file_input, outputs=image_output) with gr.Tab("Interactive Env"): gr.Markdown("### Test Classification Environment") with gr.Row(): diff = gr.Radio( ["easy", "medium", "hard"], value="easy", label="Difficulty" ) btn_create = gr.Button("Create Environment", variant="primary") status = gr.Textbox(label="Status") with gr.Row(): doc_content = gr.Textbox(label="Document", lines=4) explain_box = gr.Textbox(label="Model Reasoning", lines=4) with gr.Row(): category = gr.Number(label="Category Index", value=0, precision=0) btn_classify = gr.Button("Classify") result = gr.Textbox(label="Result", lines=2) btn_create.click( create_env, inputs=diff, outputs=[status, doc_content, explain_box, result] ) btn_classify.click( classify_doc, inputs=category, outputs=[result, doc_content, explain_box, result], ) with gr.Tab("Performance"): gr.Markdown("### Model Performance") btn_eval = gr.Button("Run Evaluation", variant="primary") score_table = gr.Dataframe( headers=["Task", "Score", "Accuracy", "Status"], label="Results" ) btn_eval.click(run_evaluation, outputs=score_table) with gr.Tab("Info"): gr.Markdown(""" ## API Endpoints (port 7861) - POST /api/reset - POST /api/step - POST /api/agent/process - POST /api/rl/train - GET /api/rl/status ## Run python app_v3.py Then open http://localhost:7860 """) if __name__ == "__main__": t = threading.Thread(target=run_flask, daemon=True) t.start() demo.launch(server_name="0.0.0.0", server_port=7860)