My dad worked as a solutions architect for enterprise companies for over 20 years, and I worked on vendor integrations myself at companies like Robinhood. So I got to see a lot of enterprise integrations products, and noticed that they all fall into two buckets. They’re either easy to use but not flexible enough for complex use cases, or flexible enough to support complex enterprise use cases but require extensive training just to learn how to use. As a result, developers have always had to make the tradeoff between: - Invest in a enterprise iPaaS that covers checks all boxes, but requires expensive training and where the vendor will raises prices immediately after you’re locked in to their ecosystem. - Adopt a simple mid-market iPaaS that solves the immediate problem, but you know you’ll have to migrate off eventually as your needs become more complex. - Dedicate resources to building integrations in-house, which takes away resources that could be put towards building features that actually get you new customers. We’re hoping to eliminate this tradeoff with an iPaaS that allows developers to build new integrations fast, but is also flexible and robust enough to scale with the company’s needs. We’ll do this with a few key features: * Declarative connector engine: Rather than taking the standard approach of building out-of-the-box connectors, we want to focus on making it as fast as possible for developers to build custom connectors while abstracting away the network-level features like backoff and error handling and every company needs to ship an integration to production. * Bundle connectors + orchestration + monitoring: Many companies have to mix and match different tools to build end-to-end integrations. They need a vendor for connectors (Airbyte, FiveTran), one for orchestration (Dagster, Airflow), and another for monitoring (Datadog). This gets expensive fast. By focusing on custom connector development rather than building our own, we free up resources to also include orchestration and monitoring features that are good enough for most use cases. * Generative AI bridge: LLMs are great at writing and explaining basic code. We can leverage that to save developers time, while also making it accessible to non developers. We’re still in alpha and the project isn’t ready for production use yet, but we wanted to launch early and get the community’s input. |