.DkNUKaKx_ZLAT4m.webp?dpl=dpl_6H4ZUkxXf57QfF3fVHzFqcbYPsVn){kind=small}
Blog / Architecture and integration
How we build the platform — connectors, workflows, MCP, and the internals of enterprise integration.
Architecture and integrationIntegration adoption: proving it cuts churnEveryone agrees customers who integrate stay longer, and nobody can show it. How to track integration adoption per customer and tie it to churn.
Architecture and integrationData integration process: where pipelines breakSeven steps move data between systems, and most failures sit between them. Where data integration pipelines break, and what to fix first.
Architecture and integrationIntegration versioning without breaking customersA new required setting goes out, and every customer who set up before it existed starts failing that afternoon. How to version integration changes safely.
Architecture and integrationIntegration errors: why support hears firstA customer's sync breaks on Tuesday, and they find out from a wrong report the next week. Why integration error handling fails customers, and what to fix.
Architecture and integrationiPaaS security: what your review should ask an integration platformAn integration platform holds credentials to every system it connects. The questions a security review should ask, what stays yours, and Tray.ai's answers.
Architecture and integrationWhy your integration backlog leads to ungoverned vibe-coded appsAn integration request sits in the queue for months. The team waiting builds it themselves, and it comes back to IT as an app nobody reviewed.
Architecture and integrationData quality monitoring: who sees bad data firstThe dashboard was wrong for three weeks and the alert went to a muted channel. Why bad data reaches the dashboard first, and what to check.
Architecture and integrationField mapping: why customer setups never finishThe customer's custom field isn't in the list, so they file a ticket and never finish setup. Why field mapping stalls customer integrations, and what to fix.
Architecture and integrationMigrating off Boomi: how big the job is, and how to make it smallerA move off Boomi is a rebuild, and its size depends on what's in your processes. How to size it, map it, and do it in waves.
Architecture and integrationSchema changes: why one renamed field breaks fiveA renamed picklist value fails nothing and changes every report that groups by it. Why schema changes break integrations, and what to fix first.
Architecture and integrationIntegration authentication: why syncs stopA customer's connection expires on a Saturday and the sync stops without a word. Why customer integrations fail at sign-in and after it, and what to fix first.
Architecture and integrationSFTP to API: why file feeds outlive migrationsA nightly file has run for eleven years and nobody knows its rules. Why batch file feeds outlive migrations, and how to move them onto APIs.
Architecture and integrationHow to go headless without forking your teamGoing headless moves integration into the AI IDE. When code and the visual builder edit different copies of a workflow, your team forks. Here's the fix.
Architecture and integrationIntegration requests: deciding what to build nextThe loudest deal sets the roadmap while the integration 300 customers asked for waits. How to track integration requests and decide what to build next.
Architecture and integrationReverse ETL vs ETL: what changes, and for whomETL loads the warehouse and reverse ETL sends its answers back out. How triggers, workloads, destinations and builders differ, and when one platform fits.
Architecture and integrationChange data capture: why CDC falls behindA nightly query on a modified date never sees a deleted row. Why change data capture pipelines fall behind production, and what to fix first.
Architecture and integration5 questions every enterprise should ask before choosing an iPaaSCan your integration layer keep up with your AI roadmap, or is it the thing slowing it down? What to ask a vendor, and which answers should worry you.
Architecture and integrationAI is exposing the limits of legacy data pipelinesAI agents are exposing the limits of legacy ETL and data pipelines. Learn why scaling AI requires unifying data transformation and orchestration.
Architecture and integration5 reasons your API Gateway won't cut it as your AI GatewaySnapchat's cloud cost per daily user jumped from $0.59 to $0.85 once AI landed. Your API gateway cannot see that, let alone do anything about it.
Architecture and integrationHas PowerShell lost its power for IT operations?PowerShell is a fine place to start and a hard place to scale. Where script-based IT automation breaks down, and what modern IT teams do instead.
Michael Douglas
Architecture and integrationReduce your integration technical debt with a modern iPaaSCIOs divert 10-20% of the IT budget to technical debt. How a modern iPaaS cuts integration maintenance so IT can get back to building new things.
Michael Douglas
Architecture and integrationWorking with the Intercom API: What you need to knowTired of manually exporting Intercom data? We’ll show you the fundamentals of the Intercom API, key data types, and how to boost retention and revenue.
Architecture and integrationIntroduction to working with Slack programmaticallyWhat Slack's API can do, how permissions and scopes work, and the common patterns for alerts, approvals, and provisioning built on top of it.
Architecture and integrationWhat is API integration? A practical guideAPI integration is how your systems share data automatically. What an API is, the common integration patterns, and what building one actually costs.
Architecture and integrationTurning real-time event streams into game-changing valueAutomate streaming event data with ease on the Tray Platform.
Architecture and integrationTips and tricks for optimizing your task usageLearn best practices around optimizing your task usage for maximum efficiency.
Architecture and integrationPersistent data storage at the solution instance levelSharing dynamic data amongst a group of workflows that are unique to a specific end-user is easy with Tray Embedded.
Architecture and integrationCallable Lookup: Employee DirectoryThe callable lookup pattern, built as an employee directory you can query by any system identifier. How it works and how to try it on your own data.
Architecture and integrationWorking with ADP Workforce Now APIs for HR and payrollA walkthrough of ADP Workforce Now's Workers v2 and Pay data input APIs, the application types you need, and how to connect payroll to the rest of HR.
Architecture and integrationSteal this custom Salesforce SyncThe real-time and backfill architecture behind a custom Salesforce sync we built to Iterable in under two weeks. Copy the build for your own CRM.
Architecture and integrationiPaaS vs RPA: What’s the difference?RPA copies mouse clicks. iPaaS calls APIs. Which one fits depends on whether your process runs on legacy screens or on systems with a real interface.
Architecture and integrationReal-time alerts for workflow errorsLearn how to setup a basic workflow alert via Slack for your integration and automation workflows on Tray.
Architecture and integrationReducing error fatigue with an error reporting dashboardIn this post we cover how to send workflow error data to your warehouse and build a heads up dashboard to monitor and fix workflow errors with ease.
Architecture and integrationWhat are webhooks, and why do they fire twice?Webhooks tell one app that something happened in another. What they are, why they often arrive twice, and how to deliver them reliably.
Architecture and integrationFive reasons why services teams need an integration solutionStruggling to meet demand for customer integrations? See how savvy services teams are delivering integrations faster, reducing churn, boosting MRR, and more.
Architecture and integrationWhat are embedded integrations? A beginner’s guideLearn the what, why, and how of embedded integration platforms in this beginner’s guide.
Architecture and integrationPower data-driven insights with AWS automation for cloud storageLearn how to augment your favorite cloud-based data storage tools (like AWS) with automation to craft processes that pipe and analyze data faster.
Architecture and integrationTranslating programming constructs to Tray workflowsIf you can write the logic in code, you can build it on a canvas. How loops, conditionals and functions map onto the equivalent Tray workflow patterns.
Architecture and integrationServerless workflows go mainstream to meet data demandsLearn how growth-focused businesses are turning to serverless workflows to achieve their data integration and automation initiatives.
Architecture and integrationAutomation drives low-code digital transformationLearn how the surge in demand for automation and integration promotes businesses to adopt a low code digital transformation strategy.
Architecture and integrationBasic ETL on the Tray platform tutorialBuild an ETL pipeline step by step on Tray, from data mapping to batching for production. The example uses BigQuery; the patterns fit any warehouse.
Architecture and integrationHow do APIs work?In 2002 Jeff Bezos ordered every Amazon team to expose its data through an interface. That mandate is why APIs matter, and how they actually work.
Want us to write about something specific?
We read every request. Tell us what you'd read.
Thanks — someone from our team will be in touch soon.