Show HN: Sequor – dbt for API Integration

[maxgrinev]

Hey HN!

We built Sequor to solve a recurring problem: choosing between two bad options for API/app integration:

1. Proprietary black-box SaaS connectors with vendor lock-in

2. Custom scripts that are brittle, opaque, and hard to maintain

Sequor is a SQL-centric workflow framework for building reliable API integrations in modern data stacks — code-first, declarative, and version-controlled. It’s an open source alternative to SaaS connectors, giving data teams full control over their integration pipelines.

Sequor combines SQL with HTTP request handling — think dbt for API integration, but with explicit flow control:

• Iterate over input tables to make parameterized API calls

• Parse and map JSON responses back into database tables

• Use SQL for logic, YAML for flows, and Jinja/Python for dynamic behavior

• Works equally well for data extraction, reverse ETL, and iPaaS-style end-to-end automation

• Git-friendly, no drag-and-drop UIs

Our goal was to bring software engineering best practices to integration workflows — without proprietary environments or vendor lock-in.

Think of Sequor as your integration foundation, not your ceiling. We provide the proven patterns — you build the custom connectors that match your data flows and business logic.

We’d love your feedback, ideas, and contributions.

Website: https://sequor.dev — with code examples

Quickstart: https://docs.sequor.dev/getting-started/quickstart

GitHub: https://github.com/paloaltodatabases/sequor

Prebuilt integrations: https://github.com/paloaltodatabases/sequor-integrations

[danbars]

Maintaining integration code is hard, no matter how you turn this. Even with your approach, you're still going to have to update your code every time Mailchimp, for example, are updating their API. If integrations is not the business you're trying to compete in, I truly think that you should use an automation platform. There are many to choose from. Use Make or Pabbly as reasonably priced options, or even n8n self-hosted if you need internal automations only.

[maxgrinev]

Thanks for the feedback! You're absolutely right that API maintenance is inherently challenging - that's actually one of the core problems we're trying to address.

The key difference with Sequor is who maintains the integration code and how. With SaaS platforms, when Mailchimp changes their API, you're at the mercy of the vendor's timeline to update their connector. With Sequor, your team has full control - you can patch it immediately, add custom logic, or handle edge cases specific to your use case.

We've found that many data teams prefer this trade-off. They'd rather own a transparent, version-controlled integration they can debug and modify than wait weeks for a vendor fix or work around black-box limitations.

In practice, breaking API changes aren't as frequent as people fear - most mature APIs maintain backward compatibility pretty well. The bigger challenge is often the day-to-day flexibility: adding custom fields, handling edge cases, or integrating with your specific data transformations. That's where having maintainable, readable code really pays off.

That said, you're spot on that SaaS automation platforms like Make, Pabbly, or n8n are great choices for many use cases! The decision really comes down to your specific needs:

If you need plug-and-play simplicity and don't mind vendor dependency → SaaS platforms If you need custom logic, full control, or integration with your existing data stack → code-first tools like Sequor

We see Sequor filling the gap for teams that have outgrown SaaS connectors but don't want to build everything from scratch. Different tools for different needs!