Home / Features / Natural Language Queries

Natural Language Queries

Query your database using plain English. No SQL knowledge required.

💬 How It Works

Datix xAgents uses advanced AI to:

  1. Understand your natural language question
  2. Analyze your database schema
  3. Generate optimized SQL query
  4. Execute and return results
  5. Visualize data when appropriate

📊 Query Patterns

Common patterns that work great:

Simple Aggregations

How many customers do we have?

What's our total revenue for 2024?

Show me average order value by month

Time-Based Analysis

Show sales trends for the last 6 months

Compare revenue this quarter vs last quarter

What's the MoM growth rate for new customers?

Grouping & Segmentation

Top 10 customers by revenue

Sales breakdown by product category

Customer distribution by region

Filtering & Conditions

Customers who haven't ordered in 90 days

Orders over $1000 in the last month

Products with inventory below 10 units

Joins & Relationships

Show me customers with their total order value

List products that have never been sold

Which sales reps have the highest close rates?

Complex Analysis

Customer lifetime value segmentation

Cohort analysis by signup month

Product affinity analysis (what's bought together)

🎯 Best Practices

✅ Be Specific

Good: "Revenue by product for Q4 2024"
Avoid: "Show me some revenue"

✅ Use Context

Include timeframes, categories, and filters in your question

✅ Follow Up

Refine results with follow-up questions: "Now filter to USA only"

✅ Name Tables/Columns

Reference specific tables when ambiguous: "from customers table"

💡 Example Questions by Use Case

📈 Sales & Revenue

  • What's our total revenue this year?
  • Show me revenue by sales rep for Q3
  • Which products generate the most revenue?
  • What's the average deal size by region?
  • Show me the sales funnel conversion rates

👥 Customer Analytics

  • How many new customers did we get last month?
  • What's the customer retention rate?
  • Show me customer lifetime value distribution
  • Which customers are at risk of churning?
  • Customer segmentation by purchase frequency

📦 Product & Inventory

  • Which products are selling the fastest?
  • Show me slow-moving inventory
  • What's the inventory turnover ratio?
  • Products with the highest return rate
  • Product performance by category

🎯 Marketing

  • Campaign ROI by channel
  • Cost per acquisition trends
  • Email campaign performance metrics
  • Customer acquisition source breakdown
  • Marketing spend vs revenue correlation

⚙️ Operations

  • Average order fulfillment time
  • Show me bottlenecks in the process
  • Resource utilization by department
  • Cost breakdown by category
  • Efficiency metrics over time

🔄 Query Refinement

Use follow-up questions to refine results:

Example Conversation:

You: Show me top customers by revenue

Agent: [Returns table with 100+ customers]

You: Limit to top 10

Agent: [Returns filtered results]

You: Now add their order count

Agent: [Adds column with order counts]

You: Sort by order count instead

Agent: [Re-sorts the data]

📊 Getting Visualizations

Request charts explicitly or let the AI suggest:

Bar Charts

Show me revenue by product as a bar chart

Compare sales across regions visually

Line Charts

Plot revenue trends over time

Show customer growth as a line chart

Pie Charts

Market share by product category as pie chart

Revenue distribution by region

🎓 Advanced Techniques

Window Functions

Show running total of revenue by month

Calculate 3-month moving average of sales

Rank customers by revenue within each region

Percentiles & Statistics

What's the 90th percentile order value?

Show me the median customer lifetime value

Standard deviation of daily sales

Date Calculations

Customers who joined in the last 30 days

Orders placed on weekends vs weekdays

Year-over-year comparison for each month

⚠️ Common Pitfalls

Avoid These Mistakes
  • Vague questions: "Show me some data" → Be specific about what you want
  • Missing timeframes: "Total revenue" → Specify the period
  • Ambiguous terms: "active" → Define what "active" means
  • Too broad: "Everything about customers" → Break into smaller questions

🚀 Tips for Better Results

1️⃣ Start Simple

Begin with basic queries, then add complexity

2️⃣ Check SQL

Review generated SQL to understand how queries work

3️⃣ Use Templates

Save common queries as templates for reuse

4️⃣ Iterate

Refine results with follow-up questions

🔍 Troubleshooting

If Results Are Wrong
  1. Check the generated SQL query
  2. Verify table/column names match your database
  3. Rephrase your question more specifically
  4. Provide additional context or constraints
  5. Contact support if issues persist

🚀 Next Steps

Next: AI Agents →