Putting It All Together

This walkthrough builds a complete Order Management domain using every DDD pattern covered in this section. We’ll model an e-commerce order lifecycle — placing orders, processing payments, managing fulfillment, and building read models.

The Domain

Our Order Management system handles:

• Placing orders — customers submit orders with items and shipping address
• Processing payments — charge the customer via a payment gateway
• Fulfilling orders — reserve inventory and create shipping labels
• Tracking status — real-time order status for customers and internal dashboards
• Notifying external systems — send events to notification and analytics services

Step 1: Model the Aggregate

The Order aggregate is an entity stream. Its identity is the order ID, and its state is derived from the sequence of domain events.

Events in the Order aggregate:

Event	Meaning
order.placed	Customer submitted an order
order.confirmed	Inventory validated, order accepted
order.paid	Payment charged successfully
order.shipped	Shipment created and dispatched
order.cancelled	Order cancelled (before shipment)

import { createClient } from "@ironflow/node";

const ironflow = createClient({ apiKey: process.env.IRONFLOW_API_KEY });

// Place a new order — create the aggregate
async function placeOrder(orderId: string, data: OrderData) {
  await ironflow.streams.append(`order-${orderId}`, {
    entityType: "order",
    name: "order.placed",
    data: {
      orderId,
      customerId: data.customerId,
      items: data.items,
      total: data.total,
      shippingAddress: data.address,
    },
  }, { expectedVersion: 0 }); // Must be a new aggregate
}

// Transition the aggregate — with invariant enforcement
async function confirmOrder(orderId: string) {
  const { events } = await ironflow.streams.read(`order-${orderId}`);
  const info = await ironflow.streams.getInfo(`order-${orderId}`);
  const version = info?.version ?? 0;
  const state = rebuildOrderState(events);

  if (state.status !== "placed") {
    throw new Error(`Cannot confirm order in status: ${state.status}`);
  }

  await ironflow.streams.append(`order-${orderId}`, {
    entityType: "order",
    name: "order.confirmed",
    data: { confirmedAt: new Date().toISOString() },
  }, { expectedVersion: version });
}

function rebuildOrderState(events: any[]) {
  return events.reduce((state, event) => {
    switch (event.name) {
      case "order.placed":
        return { ...event.data, status: "placed" };
      case "order.confirmed":
        return { ...state, status: "confirmed" };
      case "order.paid":
        return { ...state, status: "paid" };
      case "order.shipped":
        return { ...state, status: "shipped" };
      case "order.cancelled":
        return { ...state, status: "cancelled" };
      default:
        return state;
    }
  }, { status: "unknown" });
}

Step 2: Define Command Handlers

Functions react to domain events and execute business logic. The order.placed event triggers the confirmation workflow:

import { ironflow } from "@ironflow/node";
import { NonRetryableError } from "@ironflow/core";

export const validateOrder = ironflow.createFunction(
  {
    id: "validate-order",
    triggers: [{ event: "order.placed" }],
  },
  async ({ event, step }) => {
    // Check inventory availability
    const available = await step.run("check-inventory", async () => {
      return await inventoryService.checkAvailability(event.data.items);
    });

    if (!available) {
      // Cancel the order — append to the aggregate
      await step.run("cancel-order", async () => {
        await ironflow.streams.append(`order-${event.data.orderId}`, {
          entityType: "order",
          name: "order.cancelled",
          data: { reason: "Items out of stock" },
        });
      });
      return;
    }

    // Confirm the order
    await step.run("confirm-order", async () => {
      await confirmOrder(event.data.orderId);
    });
  },
);

Step 3: Build Read Models

A managed projection builds the OrderDashboard read model — an aggregated view of all orders, updated in real time:

import { createProjection } from "@ironflow/node";

export const orderDashboard = createProjection({
  name: "order-dashboard",
  events: ["order.placed", "order.confirmed", "order.paid", "order.shipped", "order.cancelled"],
  handler: (state, event) => {
    const orderId = event.data.orderId;
    const existing = state.orders[orderId] ?? {};

    switch (event.name) {
      case "order.placed":
        return {
          ...state,
          totalOrders: state.totalOrders + 1,
          orders: {
            ...state.orders,
            [orderId]: {
              id: orderId,
              customerId: event.data.customerId,
              total: event.data.total,
              status: "placed",
              placedAt: event.timestamp,
            },
          },
        };

      case "order.confirmed":
      case "order.paid":
      case "order.shipped":
      case "order.cancelled":
        return {
          ...state,
          orders: {
            ...state.orders,
            [orderId]: {
              ...existing,
              status: event.name.split(".")[1],
            },
          },
        };

      default:
        return state;
    }
  },
  initialState: () => ({ totalOrders: 0, orders: {} }),
});

A partitioned projection gives per-customer analytics:

export const customerOrders = createProjection({
  name: "customer-orders",
  events: ["order.placed", "order.shipped"],
  partitionKey: "$.data.customerId",
  handler: (state, event) => ({
    ...state,
    orderCount: state.orderCount + (event.name === "order.placed" ? 1 : 0),
    shippedCount: state.shippedCount + (event.name === "order.shipped" ? 1 : 0),
    totalSpend: state.totalSpend + (event.data.total ?? 0),
  }),
  initialState: () => ({ orderCount: 0, shippedCount: 0, totalSpend: 0 }),
});

Step 4: Coordinate with Sagas

The fulfillment saga runs when an order is confirmed — it charges payment, reserves inventory, and creates the shipment:

import { NonRetryableError } from "@ironflow/core";

export const fulfillOrder = ironflow.createFunction(
  {
    id: "fulfill-order",
    triggers: [{ event: "order.confirmed" }],
  },
  async ({ event, step }) => {
    const orderId = event.data.orderId;

    // Step 1: Charge payment
    const payment = await step.run("charge-payment", async () => {
      return await stripe.charges.create({
        amount: event.data.total,
        customer: event.data.customerId,
      });
    });
    step.compensate("charge-payment", async () => {
      await stripe.refunds.create({ charge: payment.id });
    });

    // Record payment in aggregate
    await step.run("record-payment", async () => {
      await ironflow.streams.append(`order-${orderId}`, {
        entityType: "order",
        name: "order.paid",
        data: { paymentId: payment.id },
      });
    });

    // Step 2: Reserve inventory
    const reservation = await step.run("reserve-inventory", async () => {
      return await inventoryService.reserve(event.data.items);
    });
    step.compensate("reserve-inventory", async () => {
      await inventoryService.release(reservation.id);
    });

    // Step 3: Create shipment
    const shipment = await step.run("create-shipment", async () => {
      const result = await shippingService.createLabel(event.data.address);
      if (!result.ok) {
        throw new NonRetryableError("Shipping unavailable");
      }
      return result;
    });

    // Record shipment in aggregate
    await step.run("record-shipment", async () => {
      await ironflow.streams.append(`order-${orderId}`, {
        entityType: "order",
        name: "order.shipped",
        data: { trackingNumber: shipment.trackingNumber },
      });
    });
  },
);

Step 5: Integrate Contexts

After the order ships, publish integration events for external systems:

export const notifyShipment = ironflow.createFunction(
  {
    id: "notify-shipment",
    triggers: [{ event: "order.shipped" }],
  },
  async ({ event, step }) => {
    // Publish integration events — lean, cross-context
    // step.publish() is durable and memoized
    await step.publish("notifications.shipment", {
      orderId: event.data.orderId,
      trackingNumber: event.data.trackingNumber,
    });

    await step.publish("analytics.revenue", {
      orderId: event.data.orderId,
      total: event.data.total,
    });
  },
);

Register Everything

Wire up all functions and projections in your worker:

import { createWorker } from "@ironflow/node";

const worker = createWorker({
  functions: [validateOrder, fulfillOrder, notifyShipment],
  projections: [orderDashboard, customerOrders],
});

await worker.start();

How It All Connects

DDD Pattern	Ironflow Primitive	This Example
Aggregate	Entity Stream	order-{id} stream with placed/confirmed/paid/shipped events
Command Handler	Event-triggered Function	validate-order reacts to order.placed
Domain Events	emit() + appendToStream()	order.confirmed, order.paid, order.shipped
Read Model (CQRS)	Managed Projection	order-dashboard, customer-orders
Saga	Workflow with step.compensate()	fulfill-order with payment/inventory/shipping
Integration Events	step.publish()	notifications.shipment, analytics.revenue

Run It

See the working DDD Order Management example to run these patterns as a real application with a UI.

---

Previous pages in this section:

• Why DDD with Ironflow — concept overview and strategic DDD
• Aggregates & Entity Streams — consistency boundaries
• Commands, Events & Reactions — event-driven patterns
• CQRS with Projections — read/write separation
• Sagas & Process Managers — distributed consistency