How Custom Apps Integrate With Other Platform Features

How Integration Works

Every custom app runs on the same infrastructure as the platform's built-in apps. This means your custom code can call the same internal functions that the email broadcast system, SMS broadcast system, AI chatbot, and every other module uses. There is no special API wrapper or limited SDK. Your app's functions have direct access to the full set of platform capabilities.

When you describe a feature that involves another platform module, the AI builder recognizes it and writes the appropriate function calls. For example, if you say "send a confirmation email when a new order is created," the AI writes code that calls the same email sending function used by the email broadcast system. If you say "ask GPT to summarize the customer feedback," it calls the same AI query function used by the chatbot module.

This integration happens at the code level, not through external API calls. Your custom app's functions run in the same process as the platform core, so there is no network overhead, no rate limiting between modules, and no authentication needed for internal calls. Everything works as if it were one unified application, because it is.

Email Integration

Your custom app can send emails using the same infrastructure that powers the email broadcast system. This includes both individual transactional emails (like order confirmations or password resets) and bulk email operations.

The AI can configure your app to send emails in several ways:

• Transactional emails triggered by specific events, such as a new record being created, a status change, or a user action through the API
• Scheduled email reports sent by a background job at a regular interval, like a daily summary or weekly digest
• Notification emails sent to the account owner or specific addresses when certain conditions are met
• Bulk emails to contact lists stored in the broadcast data system, using the same delivery infrastructure as dedicated email campaigns

Emails sent through your custom app use the same sending infrastructure, delivery tracking, and bounce handling as the email broadcast module. The AI writes the email content and recipient logic into your app's functions, and you can customize templates, subject lines, and sending conditions by describing what you want in the builder conversation.

SMS Integration

Custom apps can send SMS messages to any phone number, using the same messaging system that powers the SMS broadcast module. This is useful for apps that need to send alerts, reminders, verification codes, or status updates directly to users' phones.

Common SMS integration patterns include:

• Appointment reminders sent a configurable number of hours before a scheduled event
• Status notifications that text a customer when their order ships, their ticket is resolved, or their request is approved
• Two-way SMS where your app receives incoming text messages through a webhook and responds with relevant information
• Alert messages sent to the account owner when something important happens, like a high-value order or a system threshold being reached

The AI builder understands SMS message length limits, sending windows, and carrier requirements. When you describe an SMS feature, it writes code that handles formatting, timing, and delivery through the platform's established messaging infrastructure.

AI Model Integration

One of the most powerful integration capabilities is direct access to AI models. Your custom app can send queries to GPT, Claude, and other supported models, using the same functions that power the AI chatbot module.

AI integration opens up a wide range of possibilities:

• Content generation where your app creates descriptions, summaries, email copy, or any other text based on your data
• Data analysis where the AI reviews records, identifies patterns, scores leads, or categorizes incoming information
• Decision making where the AI evaluates conditions and recommends or takes actions based on its analysis
• Conversational features where your app provides a chat interface powered by AI, with custom system prompts and knowledge relevant to your use case
• Embeddings and knowledge retrieval where your app stores information as vector embeddings and uses semantic search to find relevant content when answering questions

The AI builder selects the appropriate model for each task based on what you describe. Quick classification tasks might use a fast, inexpensive model, while complex analysis or content generation might use a more capable model. You can also specify which model you prefer if you have a preference. All AI usage is billed at the standard per-model rates listed in the AI Models Guide.

Database Access Across Apps

Custom apps are not limited to reading and writing their own data. Your app can access any database table on the platform, which means it can read data created by other apps, write records that other apps consume, and build features that bridge multiple modules.

For example, a custom reporting app could read contact records from the broadcast data table, conversation history from the conversation data table, and usage logs from the reports table, then compile everything into a unified dashboard. A custom CRM could write contact records that the email broadcast system picks up for campaigns. A custom onboarding app could create chatbot configurations and knowledge base entries that the chatbot module uses immediately.

All database access follows the platform's partition key isolation model. Your app reads and writes data under your account ID, so there is no risk of accidentally accessing another user's data. The AI builder understands the database schema for every table and writes proper queries with the correct partition keys, sort keys, and field structures. For more details on database operations, see How Custom Apps Access Databases and Store Data.

Webhooks and External Services

Custom apps can receive incoming data from external services through webhook handlers. A webhook is a URL that external systems call to send data to your app, like a payment processor notifying you of a completed transaction, a form service sending a new submission, or a third-party API pushing an update.

The AI builder can set up webhook endpoints for your custom app that:

• Accept incoming POST requests from any external service
• Parse the incoming data (JSON, form data, or other formats)
• Validate the request to ensure it comes from a legitimate source
• Process the data by creating records, triggering notifications, updating statuses, or any other logic your app needs
• Return a response to the external service confirming receipt

Your app can also make outbound HTTP requests to external APIs. A background job might call a third-party service to fetch data, a webhook handler might forward processed data to another system, or an API command might enrich a record with information from an external source. The AI writes the HTTP request code, handles response parsing, and adds error handling for network failures.

Workflow and Chain Command Integration

Custom apps can participate in the platform's workflow automation system. Chain commands let you connect multiple actions into a sequence where each step can pass data to the next, and custom app commands can be both triggers and actions in these chains.

This means your custom app can:

• Trigger a workflow when something happens in your app, like a new record being created or a status changing to a specific value
• Be called by a workflow as one step in a larger automation sequence, receiving data from the previous step and passing results to the next
• Coordinate with other apps by chaining commands from the chatbot, email, SMS, or any other module together with your custom logic

For example, a customer support workflow might start when a webhook receives a new ticket, use your custom app to classify and assign it, send an SMS notification to the assigned agent, and create a follow-up reminder as a background job. Each step is a different platform feature, all coordinated through the workflow system. Learn more about workflow automation at Workflow Automation.