Data EngineeringShipped 2026Active Pipeline

Modern Data Stack & Marketing Intelligence Engine

End-to-end analytics engineering pipeline combining Cloudflare Workers, Google BigQuery, dbt Core dimensional models, and an Express analytical API.

My contribution

Data & Backend Systems Architect. Built the serverless edge ingestion pipeline, dbt Core dimensional data warehouse, data quality test suites, and the Express/Redis analytical serving API.

Project Screenshots

1 Screenshot

Data Pipeline Architecture

Modular Kimball dimensional modeling tracing raw ad events to daily facts and executive data marts.

Explore the technical implementation

A modern analytics engineering and marketing intelligence platform engineered for cross-channel attribution. Features a lightweight Cloudflare Worker (cf-bigquery-sync) capturing streaming webhooks from HubSpot CRM and Supabase with batch flushing into Google Cloud BigQuery. Includes a modular dbt Core data warehouse pipeline (dbt_marketing_intel) with Kimball dimensional modeling across staging, facts (ad performance, pacing, conversions), and daily executive data marts with automated dbt_expectations regression tests. Exposes sub-50ms analytical endpoints via an Express/TypeScript API backed by Redis caching and BullMQ queue workers.

The Engineering Challenge

Growth marketing teams faced 48-hour reporting delays and high discrepancy rates compiling ad spend and conversion metrics manually across Meta, Google Ads, LinkedIn, and CRM sources.

Engineered Solution

Engineered a serverless edge ingestion worker flushing batched JSON payloads into Google Cloud BigQuery. Built a modular dbt Core dimensional data warehouse pipeline transforming raw ad streams into daily analytical marts with strict dbt_expectations test assertions, served through a Redis-cached Express REST API.

Delivered System Outcome

Eliminated manual CSV reporting; reduced multi-channel marketing intelligence latency from 48 hours to near-real-time with automated ad-spend anomaly detection.

Key Architectural Capabilities

Cloudflare Worker serverless edge ingestion batching CRM and transactional events to BigQuery
Kimball dimensional data warehouse modeling in dbt Core (staging, daily facts, executive marts)
Data quality and statistical anomaly test suites powered by dbt_expectations
Express and TypeScript analytical microservice with sub-50ms Redis cache layers
Asynchronous sync jobs and cache invalidation managed via BullMQ and Redis

Core Stack & Infrastructure

dbt CoreGoogle Cloud BigQueryCloudflare WorkersTypeScriptNode.jsExpressPostgreSQLRedisBullMQSQL

Looking to build a similar platform?

Available for high-throughput API architecture, enterprise Next.js/Nuxt web platforms, and tailored business software.