diff --git a/examples/README.md b/examples/README.md
index 45b23ef..fc76960 100644
--- a/examples/README.md
+++ b/examples/README.md
@@ -7,6 +7,7 @@ They are public-safe product examples, not proprietary domain packs. Use them to
 ## Example sets
 
 - [Hospitality](hospitality/README.md)
+- [Postgres operational dashboard](postgres-operational-dashboard/README.md)
 - [Energy retail](energy-retail/README.md)
 - [Ecommerce](ecommerce/README.md)
 - [SaaS embedded analytics](saas/README.md)
diff --git a/examples/hospitality/README.md b/examples/hospitality/README.md
index bed7808..55da512 100644
--- a/examples/hospitality/README.md
+++ b/examples/hospitality/README.md
@@ -33,6 +33,85 @@ Hospitality reporting is most useful when it becomes a business health check, no
 - combo opportunity table
 - PMS occupancy and ADR trend
 
+## Calculation patterns
+
+### Revenue health
+
+Question:
+
+```text
+Is revenue healthy this week compared with last week?
+```
+
+Expected calculation:
+
+```text
+current_period_net_sales
+previous_period_net_sales
+absolute_change = current_period_net_sales - previous_period_net_sales
+percentage_change = absolute_change / previous_period_net_sales
+```
+
+Answer should include the selected business date range, location filter, net/gross basis, and whether refunds and voids are excluded.
+
+### Product mix and margin
+
+Question:
+
+```text
+Which categories are driving revenue but hurting margin?
+```
+
+Expected calculation:
+
+```text
+net_sales by category
+cost by category
+gross_margin = net_sales - cost
+gross_margin_percentage = gross_margin / net_sales
+```
+
+Dashboard output should be a ranked table with category, sales, margin, margin percentage, and contribution percentage.
+
+### Wastage risk for ready-made food
+
+Question:
+
+```text
+Which ready-made items are slow moving and likely to create waste?
+```
+
+Expected calculation:
+
+```text
+units_produced
+units_sold
+unsold_units = units_produced - units_sold
+sell_through_percentage = units_sold / units_produced
+estimated_waste_value = unsold_units * unit_cost
+```
+
+If production or waste fields are missing, IntraQ should say which fields are required instead of guessing.
+
+### Basket and combo opportunity
+
+Question:
+
+```text
+Which products are frequently bought together?
+```
+
+Expected calculation:
+
+```text
+orders containing product_a and product_b
+pair_frequency
+pair_revenue
+attach_rate = orders_with_pair / orders_with_product_a
+```
+
+Dashboard output should show product pair, pair frequency, attach rate, pair revenue, and suggested promotion angle.
+
 ## Trust checks
 
 - Confirm the date basis: transaction date, business date, stay date, or invoice date.
diff --git a/examples/postgres-operational-dashboard/README.md b/examples/postgres-operational-dashboard/README.md
new file mode 100644
index 0000000..2f076b8
--- /dev/null
+++ b/examples/postgres-operational-dashboard/README.md
@@ -0,0 +1,110 @@
+# Postgres operational dashboard example
+
+This example shows how to position IntraQ for a team with operational data already stored in PostgreSQL.
+
+The goal is not only to connect Postgres. The goal is to turn SQL-backed tables into trusted AI answers and reusable dashboard components.
+
+## Example source tables
+
+```text
+orders
+  id
+  business_date
+  location_id
+  channel
+  gross_sales
+  discount_amount
+  refund_amount
+  tax_amount
+  net_sales
+  cost_amount
+
+order_items
+  id
+  order_id
+  product_id
+  quantity
+  gross_sales
+  discount_amount
+  net_sales
+  cost_amount
+
+products
+  id
+  name
+  category
+  sku
+
+locations
+  id
+  name
+  region
+```
+
+## Metadata IntraQ needs
+
+- Date field: `orders.business_date`
+- Revenue metric: `sum(orders.net_sales)`
+- Gross sales metric: `sum(orders.gross_sales)`
+- Discount metric: `sum(orders.discount_amount)`
+- Refund metric: `sum(orders.refund_amount)`
+- Margin metric: `sum(orders.net_sales - orders.cost_amount)`
+- Relationships:
+  - `orders.id = order_items.order_id`
+  - `order_items.product_id = products.id`
+  - `orders.location_id = locations.id`
+
+## Questions to try
+
+```text
+How is revenue trending by day for this month?
+```
+
+Expected output:
+
+- line chart
+- grouped by `business_date`
+- metric: net sales
+- evidence: SQL query and row count
+
+```text
+Which products have high revenue but low margin?
+```
+
+Expected output:
+
+- table sorted by net sales descending
+- columns: product, category, net sales, cost, gross margin, gross margin percentage
+- warning if cost fields are missing
+
+```text
+Which locations are underperforming compared with last week?
+```
+
+Expected output:
+
+- table by location
+- current period net sales
+- previous period net sales
+- absolute change
+- percentage change
+- selected date range and comparison range
+
+```text
+Create a dashboard with revenue trend, revenue by channel, margin by category, and top products.
+```
+
+Expected output:
+
+- four dashboard components
+- each component has a saved data source/table binding
+- each chart remains backed by SQL, not static data
+
+## Trust checks
+
+- Always confirm the selected date range.
+- Confirm whether revenue uses gross, net, tax-inclusive, or tax-exclusive sales.
+- Confirm how discounts, refunds, voids, and cancelled orders are handled.
+- Confirm whether results are filtered by location, tenant, company, or all accessible data.
+- If a relationship or metric definition is missing, ask for it before generating a final answer.
+