superglue Is Taking Aim at the Brittle SQL Scripts and Cron Jobs Holding Enterprises Together
Superglue, a Winter 2025 Y Combinator company, is attacking one of the least glamorous and most fragile layers of the modern enterprise stack: the brittle SQL scripts, cron jobs, and one-off connectors that quietly run most day-to-day operations. Founded in 2024 by Adina Goerres and Stefan Faistenauer, the four-person team is building an open-source platform that replaces those hand-rolled data pipelines with clean, versioned, trustworthy syncs, so migrating systems, adding tools, or scaling processes stops being a guaranteed source of outages.
Founded in 2024 and based in San Francisco, the category superglue is entering, enterprise integration and data plumbing, is deceptively enormous. Almost every company above a certain size runs on a web of Airflow DAGs, scheduled SQL scripts, API glue, and internal tools that were written years ago by engineers who have since moved on. The code is mostly undocumented, the logic is mostly tribal knowledge, and the consequences of a single broken connector can be anything from a failed payroll run to a missed board report. When a new CRM is rolled out, or an ERP is migrated, those breakages multiply. This is the kind of problem that is widely acknowledged but almost never solved. Incumbents in the iPaaS and data-integration world have been large and profitable for years, and the problem is still growing. What is new is that modern AI is finally capable of reading existing data pipelines, reasoning about their intent, and reconstructing them as clean, auditable syncs. That is superglue’s core wedge.
The product’s framing is sharp and honest. Rather than pitching AI as an additive layer, superglue is positioned as a replacement for the integrations that already exist, rebuilding the logic behind them so they are understandable, versioned, and maintainable. That is a materially different value proposition from “we have an AI feature for your data pipeline”: it is “your data pipeline was the problem, and now it isn’t.” Customers running on superglue ship AI projects faster precisely because the underlying connective tissue is finally trustworthy, and they spend less on maintenance because the same layer that builds the integrations is also the layer that keeps them correct. The open-source posture at github.com/superglue-ai/superglue is the right distribution strategy for this category; enterprise data engineers are an audience that tends to trust code they can read, extend, and self-host.
The founders bring the right mix for this category. Adina Goerres previously worked on GTM at Scalable Capital and on commercial projects at Farfetch, with a master’s from TUM and CDTM, a background that matters for a product whose sales motion requires speaking to both engineering leaders and business-side buyers. Stefan Faistenauer is building the open-source engine; he was previously a software engineer and a Google APM, with a master’s from Stanford. The combination of a commercially minded founder and a technically deep one is a reliable shape for infrastructure companies that ultimately need to sell into the enterprise while maintaining credibility with developers. Superglue’s focus areas span open source, API, enterprise, enterprise software, and AI, the union of exactly the right categories for what it is building.
The opportunity in enterprise integration is well-understood, a multi-ten-billion-dollar market that has produced durable public companies, but the entry point has historically been tough because incumbents are embedded. AI is the credible disruption vector here: it changes the unit economics of building and maintaining integrations so dramatically that a well-executed new entrant can actually displace existing gluecode, rather than merely sitting alongside it. Superglue is betting that the combination of an open-source core, AI-native rebuilds, and a clean developer experience will be enough to win that wedge. With four people and a Winter 2025 YC stamp, the team is still early, but the strategic positioning is one of the sharper ones in the current enterprise AI wave.