Work that made a measurable difference.

The challenge

The revenue team had no reliable view of pipeline health. Data lived in five disconnected systems — CRM, product, billing, support, and marketing — and every weekly report required manual reconciliation in spreadsheets.

The work

Designed and built a unified data model in Snowflake pulling from all five source systems via Fivetran. Developed dbt transformations to produce a clean, tested layer for GTM analytics. Built Looker dashboards for pipeline coverage, stage conversion, and rep performance.

The challenge

Customer success managers were working from gut feel. There was no systematic way to identify at-risk accounts before they churned, and the team was too large to give every account equal attention.

The work

Built a customer health scoring model using product usage, support ticket history, billing signals, and engagement data. Developed a churn prediction model in Python (scikit-learn) trained on 18 months of historical data. Delivered scores into the CRM so CSMs could prioritize their book of business.

The challenge

The company had grown through acquisition and inherited three separate data stacks. Leadership needed an independent view of what to consolidate, what to retire, and what to build — before committing to a multi-year platform investment.

The work

Conducted a structured assessment of all three stacks: data sources, transformation layers, BI tooling, and team workflows. Interviewed stakeholders across engineering, analytics, finance, and product. Delivered a prioritized roadmap with build-vs-buy recommendations and a migration sequencing plan.