User Prompts

# User Prompts for Obviously AI

## Prompt 1: Sales Prediction
"I uploaded our sales data with columns: deal_id, deal_size, industry, company_size, lead_source, sales_rep, days_in_pipeline, number_of_meetings, proposal_sent (yes/no), competitor_involved (yes/no), deal_won (yes/no). Predict which open deals will close: (1) Build a model predicting deal_won, (2) Rank our current 50 open deals by win probability, (3) Tell me the 3 most important factors that determine whether a deal closes, (4) For our top 10 highest-value open deals, show individual predictions with reasoning, (5) Suggest what our sales team should focus on to improve win rates, (6) Estimate the total pipeline value we're likely to close this quarter."

## Prompt 2: Customer Lifetime Value Prediction
"Here's our customer data: customer_id, acquisition_date, acquisition_channel, first_purchase_amount, total_orders, total_revenue, average_order_value, days_between_orders, product_categories_purchased, returns_count, support_interactions, referrals_made, is_subscribed (yes/no). Predict customer lifetime value: (1) Build a model to predict total_revenue at the 12-month mark, (2) Segment customers into value tiers (high, medium, low) with characteristics of each, (3) Identify which acquisition channels produce the highest-LTV customers, (4) Predict LTV for our 200 newest customers (acquired this month), (5) Show what behaviors in the first 30 days predict high LTV, (6) Calculate the ROI of increasing retention rate by 5% based on the model."

## Prompt 3: Quality Defect Prediction
"My manufacturing data includes: batch_id, production_line, operator_id, raw_material_supplier, temperature_setting, pressure_setting, humidity_level, machine_speed, shift (morning/afternoon/night), machine_age_months, last_maintenance_days_ago, defect_found (yes/no), defect_type. Predict quality defects: (1) Build a model to predict defect_found, (2) Show which production parameters most influence defect rates, (3) Find the optimal settings for temperature, pressure, and speed to minimize defects, (4) Identify which operators and production lines have highest defect rates (controlling for other factors), (5) Create a real-time prediction system that flags high-risk batches before they complete, (6) Estimate the cost savings from reducing defects by the amount this model would enable."

## Prompt 4: Employee Performance Prediction
"I have employee performance reviews data: employee_id, department, role_level (junior/mid/senior), years_experience, education_level, training_hours_completed, projects_completed, peer_review_score, manager_rating, on_time_delivery_percentage, innovation_score, collaboration_score, performance_rating (1-5). Predict performance rating: (1) Build a model to predict next quarter's performance rating, (2) Identify the factors that most strongly predict high performance, (3) Find employees whose predicted performance is much higher or lower than their current rating (hidden gems and at-risk performers), (4) Analyze whether training hours actually correlate with better performance, (5) Recommend a data-driven approach to employee development based on the patterns, (6) Show how the predictors differ across departments."

## Prompt 5: Real Estate Price Prediction
"My property dataset includes: property_id, location_zip, property_type (house/condo/townhouse), square_footage, bedrooms, bathrooms, year_built, lot_size, garage_spaces, has_pool, has_basement, school_rating_nearby, crime_rate_area, distance_to_downtown_miles, last_sold_price, last_sold_date, current_listing_price. Predict property value: (1) Build a model to predict current_listing_price, (2) Show which features add the most value (what's a pool worth? an extra bedroom?), (3) Identify undervalued properties (predicted price significantly above listing price), (4) Create a price estimator where I can input property features and get a predicted price range, (5) Show how prices vary by location, property type, and age, (6) Predict which zip codes will see the most appreciation based on current trends."