Case studies

Before and after, in real numbers.

Two projects from very different contexts. Same discipline behind them.

01 / Telehealth (US)

Leafwell

The problem

Duplicated metrics and BI logic drift across dashboards used by executives, operations and product. Every team brought a different number to the same question.

The approach

Unified dbt models on Redshift, orchestration with Airflow, and governance workflows for A/B tests and cohort analysis. Ad-hoc analysis in Python, with Claude used to speed up model review and documentation.

The outcome

Aligned KPI definitions across teams, integrity across reporting surfaces, and an ongoing partnership with platform and engineering.

100%

KPIs unified

5+

Aligned teams

30+

States served

2+ yrs

Active partnership

dbtAirflowRedshiftMetabasePythonClaude

02 / Gaming community

Tales Inc.

The problem

A platform with no analytics stack. Product decisions made on intuition. No retention, no cohorts, no abuse detection.

The approach

Event tracking, metadata ingestion, Metabase dashboards, retention modeling, REST API integration for automated governance, and abusive-behavior detection pipelines.

The outcome

Cohort-driven product decisions. Sustainable growth. Automated abuse detection cut weekly moderation work by hours.

125K

Registered users

20K+

Active members

300%

ARR growth

500%

DAU growth

MetabasePostgresPythonREST APIsdbt

Your project could be the next case study.