May 2022 Summaries

Synchronous integration is a method used in B2B SaaS where data from a third-party app is needed immediately, requiring a webhook URL call and a prompt response to confirm data processing before proceeding with subsequent tasks. This process is exemplified by an integration involving the Acme app, QuickBooks, and Amazon SES, where an order request triggers a series of actions to check inventory, create invoices, and notify users of order statuses. The integration is designed to update the Acme app with a status confirmation, distinguishing it as synchronous, though it could also be adapted to run asynchronously if needed. The choice between synchronous and asynchronous integration depends on specific business requirements, and Prismatic's integration designer offers tools to simplify these processes, allowing for easy transitions between synchronous and asynchronous setups as necessary.

Webhooks are essential for SaaS integrations as they enable fast and simple data transfer via HTTP callbacks between systems, triggered by data changes. To ensure secure data transfer and prevent spoofing, webhooks should utilize HMAC, or hashed message authentication code, which uses both a secret cryptographic key and an HTTP request body to generate a hash that verifies the legitimacy of data sent to webhook endpoints. By hashing the payload with HMAC using a secret key and including the resulting hash in the HTTP request header, the destination app can authenticate the data by comparing this hash with its own computed hash. This method is widely supported across various programming languages and is used by popular applications like Slack, Dropbox, and Shopify to secure their webhooks. HMAC ensures data integrity without requiring deep encryption knowledge, making it a practical choice for protecting data in SaaS integrations.

Selecting an embedded Integration Platform as a Service (iPaaS) is a crucial decision for companies looking to enhance app functionality and customer retention through efficient integration deployment without increasing developer workload. An embedded iPaaS enables businesses to quickly adapt to changing customer needs by facilitating new integrations and allowing customers to manage these integrations independently, thereby reducing the burden on customer-facing teams. It is important for the chosen iPaaS to possess end-to-end functionality, including a low-code integration designer, code-native building experience, built-in and custom connectors, an integration marketplace, and deployment and support tools. Evaluating an iPaaS involves engaging the right stakeholders, confirming the platform’s scalability and functionality in a test environment, and conducting a realistic proof of concept to ensure it meets real-world integration complexities. The iPaaS should seamlessly integrate with existing tools and processes to be effectively adopted. By providing a strategic advantage, the right embedded iPaaS can significantly improve integration delivery, enhance customer satisfaction, and support long-term business goals.

Scheduled integrations are crucial for transferring data from systems without real-time capabilities, such as webhooks or pub/sub systems, by running at predetermined intervals. Unlike event-driven integrations that react to data changes, scheduled integrations operate on a pull mechanism, querying data at set times, which, while not real-time, can still be timely. They are particularly useful for interacting with APIs that only support queries or for systems that export data as files, common in non-SaaS or legacy environments. The process involves setting up the integration to check source systems at optimal times for data availability, employing custom components to handle data retrieval and processing, and leveraging built-in components from platforms like Prismatic to facilitate file transfers and API interactions. An example discussed involves a scheduled file import from Dropbox into an app, demonstrating a loop that processes each file and updates orders in the Acme app, showcasing the adaptability of scheduled integrations for various data types and formats. Despite the shift towards event-driven APIs in newer apps, scheduled integrations remain a vital solution for integrating with legacy systems, offering a methodical approach to data exchange in environments lacking real-time data propagation capabilities.