import os import logging from datetime import datetime import pandas as pd from dotenv import load_dotenv from simple_salesforce import Salesforce from fastapi import FastAPI from fastapi.middleware.cors import CORSMiddleware import gradio as gr from transformers import pipeline # Setup logging logging.basicConfig(level=logging.INFO) logger = logging.getLogger(__name__) # Load environment variables load_dotenv() SF_USERNAME = os.getenv("SF_USERNAME") SF_PASSWORD = os.getenv("SF_PASSWORD") SF_SECURITY_TOKEN = os.getenv("SF_SECURITY_TOKEN") SF_DOMAIN = os.getenv("SF_DOMAIN", "login") # Hugging Face sentiment analysis model sentiment_pipeline = pipeline("sentiment-analysis") # FastAPI app app = FastAPI() app.add_middleware( CORSMiddleware, allow_origins=["*"], allow_credentials=True, allow_methods=["*"], allow_headers=["*"], ) # Connect to Salesforce def connect_to_salesforce(): try: sf = Salesforce( username=SF_USERNAME, password=SF_PASSWORD, security_token=SF_SECURITY_TOKEN, domain=SF_DOMAIN ) logger.info("Salesforce connected successfully.") return sf except Exception as e: logger.error(f"Salesforce login failed: {e}") return None # Lookup Vendor ID by Name (assuming Vendor_Name__c references Vendor__c) def get_vendor_id_by_name(sf, vendor_name): try: # Query the Vendor__c object (adjust object name if different, e.g., Account) query = f"SELECT Id FROM Vendor__c WHERE Name = '{vendor_name}' LIMIT 1" result = sf.query(query) if result["totalSize"] > 0: return result["records"][0]["Id"] else: logger.error(f"No vendor found with name: {vendor_name}") return None except Exception as e: logger.error(f"Error querying vendor ID for {vendor_name}: {e}") return None # Save to Salesforce def save_to_salesforce(data): sf = connect_to_salesforce() if not sf: return "Salesforce connection failed" try: vendor_name = data["vendor_name_id"] # Always look up the vendor ID by name vendor_id = get_vendor_id_by_name(sf, vendor_name) if not vendor_id: return f"Error: No vendor found with name '{vendor_name}'" result = sf.Vendor_Performance__c.create({ "Vendor_ID__c": int(data["vendor_id"]), "Vendor_Name__c": vendor_id, # Use the resolved ID "Score__c": data["score"], "Timeliness_Score__c": data["timeliness_score"], "Issue_Count__c": int(data["issue_score"] / 3), "Feedback_Rating__c": data["feedback_score"], "Evaluation_Date__c": data["evaluation_date"], "Rationale__c": data["rationale"] }) logger.info(f"Record saved to Salesforce: {result}") return "Saved successfully" except Exception as e: logger.error(f"Save failed: {str(e)}") return f"Error: {str(e)}" # Score vendor def score_vendor(vendor_id, vendor_name_id, delivery_records, issue_counts, nps_values): if not vendor_id or not vendor_name_id: return {"error": "Vendor ID and Vendor Name are required."} if not (0 <= delivery_records <= 100): return {"error": "Delivery % must be 0–100"} if not (0 <= nps_values <= 100): return {"error": "NPS must be 0–100"} timeliness_score = round(delivery_records * 0.4, 2) issue_score = round((10 - min(issue_counts, 10)) * 3, 2) feedback_score = round(nps_values * 0.3, 2) total_score = round(timeliness_score + issue_score + feedback_score, 2) eval_date = datetime.today().strftime('%Y-%m-%d') rationale = f"Timeliness: {timeliness_score} + Issues: {issue_score} + Feedback: {feedback_score}" result = { "vendor_id": vendor_id, "vendor_name_id": vendor_name_id, "score": total_score, "timeliness_score": timeliness_score, "issue_score": issue_score, "feedback_score": feedback_score, "evaluation_date": eval_date, "rationale": rationale } save_status = save_to_salesforce(result) result["status"] = save_status return result # Format for Gradio def format_output(vendor_id, vendor_name_id, delivery_records, issue_counts, nps_values): result = score_vendor(vendor_id, vendor_name_id, delivery_records, issue_counts, nps_values) if "error" in result: return result["error"], None df = pd.DataFrame([result]) return result["status"], df[[ "vendor_id", "vendor_name_id", "score", "timeliness_score", "issue_score", "feedback_score", "evaluation_date", "rationale" ]] # Batch CSV Upload def process_uploaded_file(file): try: df = pd.read_csv(file.name) required = {"vendor_id", "vendor_name_id", "delivery_records", "issue_counts", "nps_values"} if not required.issubset(df.columns): return f"Missing required columns: {', '.join(required)}", None results = [] for _, row in df.iterrows(): res = score_vendor(row["vendor_id"], row["vendor_name_id"], row["delivery_records"], row["issue_counts"], row["nps_values"]) if "error" not in res: results.append(res) if not results: return "No valid records found", None return "Batch processed", pd.DataFrame(results)[[ "vendor_id", "vendor_name_id", "score", "timeliness_score", "issue_score", "feedback_score", "evaluation_date", "rationale" ]] except Exception as e: return f"Error processing file: {e}", None # Gradio UI with gr.Blocks(title="Vendor Performance App") as demo: gr.Markdown("## 📊 Vendor Performance Scoring Tool") with gr.Tabs(): with gr.Tab("Single Vendor"): vendor_id = gr.Textbox(label="Vendor ID") vendor_name_id = gr.Textbox(label="Vendor Name") # Updated label to only expect name delivery = gr.Slider(0, 100, label="Delivery %") issues = gr.Number(label="Issue Count", minimum=0) nps = gr.Slider(0, 100, label="NPS") status = gr.Textbox(label="Status", interactive=False) result_table = gr.Dataframe(headers=[ "Vendor ID", "Vendor Name", "Score", "Timeliness", "Issues", "Feedback", "Date", "Rationale" ]) submit = gr.Button("Submit") submit.click( fn=format_output, inputs=[vendor_id, vendor_name_id, delivery, issues, nps], outputs=[status, result_table] ) with gr.Tab("Upload CSV"): file_upload = gr.File(label="Upload CSV (vendor_name_id should contain vendor names)", file_types=[".csv"]) file_status = gr.Textbox(label="Upload Status") file_result = gr.Dataframe(headers=[ "Vendor ID", "Vendor Name", "Score", "Timeliness", "Issues", "Feedback", "Date", "Rationale" ]) file_upload.change( fn=process_uploaded_file, inputs=file_upload, outputs=[file_status, file_result] ) if __name__ == "__main__": import uvicorn demo.launch(server_name="0.0.0.0", server_port=7860) uvicorn.run(app, host="0.0.0.0", port=8000)