We built what we needed after the failure
A pipeline broke in production. Millions of downstream decisions ran on stale data for three days before anyone noticed. Decube exists because that should never happen again.
Why Decube exists
In 2022, Jatin Solanki was leading data infrastructure at a Series B fintech. A Fivetran connector to their payment processor silently stopped loading at 11pm on a Wednesday. The failure did not surface until Friday afternoon, when a finance analyst noticed that the weekly revenue report looked wrong.
By then, 48 hours of dashboards, automated reports, and ML model inputs had been running on stale data. The root cause took six hours to find. A column in a source table had changed type from VARCHAR to INTEGER during a backend migration. No one was notified. No monitor caught it. The pipeline failed silently.
Jatin spent three weeks evaluating every data observability and catalog product on the market. The observability tools showed that something broke but not why or what it touched downstream. The catalog tools had rich metadata but no live quality signals. No single tool gave the full picture - what changed, what it affected, and who was responsible.
He left to build Decube with co-founder Priya Nair in early 2023. The goal was not to build another alert tool. It was to give data teams genuine context - lineage plus quality plus ownership, connected so that one alert shows the complete blast radius. Decube is backed by Iterative Accelerator through a seed round investment.
What we actually believe
Not a mission statement. The specific opinions that drive product decisions.
Lineage without quality is useless
A map that shows the path but not whether the road is open does not help. Lineage and quality are the same problem, not two separate products.
Alerts without context create noise
Every data team has a #data-alerts channel that everyone has muted. That is what happens when you alert without showing ownership and impact. We route alerts to the right person with the right context.
Tools that require setup never get used
We spent months making the first connection take under 5 minutes. A platform that requires a week of YAML configuration and a dedicated deployment engineer will not be used by the teams who need it most.
2023
Founded
Singapore
Headquarters
50+
Data source integrations
Seed
Funding stage
Small team. Deep domain expertise.
The Decube founding team has spent a combined 18 years building and operating data infrastructure - at fintechs, healthcare companies, and infrastructure startups. We have debugged the pipelines that Decube is now built to monitor.
We are a seed-stage company. Every engineer on the team talks to customers weekly. Product decisions come from watching actual debugging sessions, not from surveys.
Meet the Team
Help us build the context layer for data
We are looking for data engineers who care about correctness as much as speed. And customers who want to debug pipelines in minutes, not days.