Cosuno Scales Product Analytics and AI with Estuary Flow

Challenge

As a fast-growing SaaS platform serving construction companies, Cosuno needed deep visibility into how users were engaging with its tendering tools and new AI features. Product decisions relied heavily on usage analytics, including how subcontractors were invited, how projects were created, and how a new AI agent was used within the platform.

However, syncing this data from multiple sources, including an AWS Postgres production database, Salesforce, and Chargebee, into BigQuery for unified analytics in Looker was a growing challenge.

“With Fivetran, we were spending much more just to get data where needed. We needed something leaner that still did the job.”
— Maximilian Seifert, CTO & Co-founder, Cosuno

Cosuno’s small but mighty data team needed a solution that:

• Automated data movement from diverse sources
• Integrated cleanly with dbt and BigQuery
• Reduced tool cost without sacrificing reliability
• Freed up data analyst time from pipeline management

Solution

Cosuno adopted Estuary Flow to power its end-to-end real-time data integration. With native support for their core data sources (AWS Postgres, Salesforce, Chargebee) and seamless delivery into BigQuery, Flow eliminated the need for custom pipeline development.

Estuary’s managed connectors and change data capture (CDC) support ensured always-fresh product analytics with zero maintenance overhead. Combined with dbt transformations, this enabled Cosuno to build clean, queryable data marts used by teams across product, marketing, and analytics.

“The biggest value is that we don’t need to invest engineering effort into pipelines. Estuary makes the whole thing automatic, and that’s huge for a lean team like ours.”
— Maximilian Seifert

Flow also supports a mission-critical ML use case: powering an in-app algorithm that ranks subcontractors. Precomputed metrics are synced to BigQuery multiple times a day via Estuary and surfaced to users directly within the Cosuno app, making Flow part of a production-facing operational pipeline.

AI Use Case

One of Cosuno’s most innovative features is an AI-powered assistant that helps users interact with construction project data, including bills of quantities, through a natural language interface. This assistant enables users to request edits, flag inconsistencies, and query project specs via chat.

Another key AI-driven capability is subcontractor ranking. When users search for subcontractors, Cosuno’s platform displays ranked results using a machine learning model that factors past performance, project types, and customer fit.

These rankings are powered by precomputed metrics stored in BigQuery and refreshed continuously throughout the day using Estuary Flow and dbt. That means Estuary is not just a reporting tool — it’s powering customer-facing product features that directly impact the user experience.

“Our ranking model runs on BigQuery metrics and is updated several times daily. Estuary is part of the operational data loop that makes that possible.”
— Maximilian Seifert, CTO & Co-founder, Cosuno

Results

Cosuno’s decision to switch from Fivetran to Estuary Flow paid off immediately:

• 50% cost reduction for the same volume of data movement
• Zero pipeline incidents since deployment
• Faster iteration cycles on new AI features thanks to better product analytics.
• Efficient decision-making across product and go-to-market teams

“Our data pipelines just run, and the savings are clear. Estuary lets us focus on using our data, not managing it.”
— Maximilian Seifert, CTO & Co-founder, Cosuno

With Estuary Flow, Cosuno turned complex, multi-source data ingestion into a fully automated, self-service process, unlocking better product development and customer insights without the overhead.