How Clue Turned Iceberg Operations into a Strategic Advantage with Ryft

Company Profile

Clue (helloclue.com) is the #1 women-led period and cycle tracker, with 10M active users. The company generates revenue through subscriptions across mobile and web and is expanding into new offerings and integrations.

Clue’s data organisation, led by Director of Data Linda Cen, is responsible for building and operating the company’s data platform and enabling analytics and reporting across the business. The platform stores tens of terabytes of data in Apache Iceberg, running on AWS.

The Challenge: Iceberg at Scale with Limited Resources

After migrating from a legacy data warehouse to Iceberg to improve flexibility and cost efficiency, the platform scaled successfully. As adoption increased, operating Iceberg at scale introduced additional layers of operational complexity that required more structured approaches.

Manual operations became a tax

Iceberg table maintenance required custom pipelines for compaction, snapshot expiration, orphan file cleanup, data retention, and related tasks. As the number of tables and workloads increased, these activities became increasingly repetitive and less aligned with the team’s focus on platform evolution and business enablement.

Operational friction emerged as Iceberg usage scaled

While analytics engineers understood the data model, operational Iceberg metadata was not easily discoverable in a single place. As a result, answering day-to-day questions about data assets often required data engineering involvement, creating friction and slowing downstream work.

Balancing efficiency, cost control, and governance

As Iceberg adoption expanded, Clue prioritised running its data platform efficiently without increasing operational overhead. The team sought stronger visibility into storage usage and more consistent governance practices, particularly given the sensitivity of health-related data. Reducing outdated or inactive tables became both a cost-optimization and a risk-management priority.

Clue explored AWS native tools and observability options, but none provided a unified way to automate Iceberg operations while also offering deep visibility and a usable operational interface.

The Solution: Ryft as the Iceberg Control Plane

Clue adopted Ryft as the operational, optimisation, and governance layer for Iceberg.

Adaptive optimisation, automated

Ryft replaced custom scripts with policy-driven automation for compaction, snapshot management, orphan cleanup, and retention. Maintenance runs continuously and safely, without manual tuning or new pipelines.

Visibility and self-service by default

Ryft provides the data team with a unified view of Iceberg operations, centralising operational context and reducing the need for internal coordination when answering routine questions about data assets.

Embedded expertise and safer operations

When Clue hired a new data engineer, Ryft supported onboarding with Iceberg-specific walkthroughs and ongoing guidance. Ryft acts as a second set of eyes on changes, reducing operational risk.

Governance without overreach

As a healthcare data company, Clue is conscious about data privacy and governance. Ryft’s BYOC (bring-your-own-cloud) model aligns with that posture, enabling Clue to retain complete control of its data while automating Iceberg operations. Ryft operates on metadata only, allowing the team to gain optimisation, visibility, and enforcement without expanding access to sensitive health data or any other personal data.

Results

Operational leverage and cost control

The data team reclaimed time previously spent on maintenance and ad hoc support. Engineers now focus on onboarding new data sources, improving pipelines, and supporting business growth. Using Ryft’s visibility, Clue identified unused data assets, applied retention policies, and reduced storage footprint.

Improved performance and confidence

Compared to the previous architecture, Iceberg delivers greater flexibility and performance. With Ryft’s adaptive optimisation and observability, the team operates with greater confidence, safer rollbacks, and reduced risk of accidental changes.

Foundation for future growth

As Clue expands into new markets, use cases, and audiences with different consent and retention requirements, Ryft provides the control plane to enforce differentiated governance policies while maintaining visibility across a growing data platform.

Clue now runs an Iceberg platform that scales predictably without scaling headcount.

Business and technical outcomes

• 99% reduction in engineering time
• Two weeks to a fully managed Iceberg data lake