Cesar R.

Top Rated Plus · 100% Job Success · $600K+ earned · Microsoft AI Cloud Partner. 13+ years in data, including BNY and the U.S. Department of Education. I fix tracking and build the data layer behind it, from a single GA4 and GTM cleanup to unifying 25 sources for an eight-brand media group with an AI agent its team queries in plain language. Most marketing analysts stop at the dashboard. Most data engineers never touch the tracking. I work on both sides, so the conversion on your site and the number in the board report come from one system, not pieces that don't connect. THE PROBLEM I SOLVE You're running multiple brands, divisions, or locations, and the numbers don't tie. Ad platforms claim conversions your CRM never sees. Each brand reports a little differently. Someone rebuilds the same spreadsheet every month, and by the time leadership sees a number, nobody fully trusts it. Some clients need one tracking setup fixed this week; others need the whole data layer rebuilt. I do both, and the small jobs often show where the bigger problems are. WHAT I BUILD - Tracking and attribution: GA4, Google Tag Manager, Meta Pixel and Conversions API, Google Ads conversions. Clean conversion data you own, tied to the brand or location that earned it. - Ingestion and warehouse: automated pipelines from your CRM, ad platforms, operational software, spreadsheets and APIs into BigQuery, Snowflake, or PostgreSQL. - Modeling and metric definitions: SQL and dbt models orchestrated with Dagster, so every brand, division and report uses the same number. - Reporting: Power BI, Looker Studio, or Tableau. By brand, by division, and rolled up. - AI on your data: an AI agent your team can ask business questions in plain language, plus automated reconciliation between systems. Every automation ships with approval gates, an error queue, and an audit log. KPIs get defined once in the data layer, not inside every report. AI comes last, once the data underneath is trusted. WAYS TO START - Tracking Audit & Fix: GA4, GTM, ad pixels and conversions checked end to end, a plain-English report, and the fixes made. - Data Foundation, Phase 1: your highest-pain sources in one warehouse, KPIs defined once, and one dashboard everyone trusts. - AI Agent Pilot: one agent answering your team's questions from approved data, with read-only access and an audit trail. - Monthly Retainer: ongoing tracking upkeep, new sources, reporting and AI use cases. RECENT WORK - Multi-brand media company (8 brands): unified 25 data sources in BigQuery and built an AI agent the team queries in plain language. A multi-hour monthly reporting process now returns answers in seconds. - Media and education group (6 brands): centralized 19 sources (HubSpot, Salesforce, GA4, Google Ads, Meta, LMS and more) in BigQuery, replaced Tableau with Looker Studio, and added a Claude agent once the data was trusted. 1,200+ staff hours recovered a year. - PE-backed restoration platform (7 divisions, 3 regions): weekly board reporting automated on PostgreSQL, Dagster and Power BI. Five hours back every week; new acquisitions appear in reporting the following Tuesday. Retained for ongoing support. - Hospitality group (7 brands): found up to 8 GA4 properties cross-firing on a single site, rebuilt measurement with one clean property per brand, and tied reservations to ad spend as valued conversions. - Cattle company: Copilot Studio agent in Microsoft Teams answering staff questions live from Azure SQL, with read-only access to approved views. - Consumer products company: Snowflake warehouse and Tableau reporting, retained since 2022. - PE value-creation firm: profit-by-customer analysis reconciling 2.08M transactions to the dollar, surfacing $0.6–1.1M in annually realizable margin. WHO I WORK WITH Businesses running multiple brands, divisions, or locations (media and marketing groups, PE-backed platforms, hospitality groups, multi-location operators), the agencies and PE sponsors behind them, and growing companies that want their tracking and reporting done right the first time. HOW I WORK I lead every engagement from architecture through delivery, with a vetted team building underneath. Your data stays in your environment. Everything is documented and handed over: architecture, schema, data dictionary, metric definitions and source code. You own it. Hourly, by project, or on retainer. Tell me what systems you're running and where the numbers break, and I'll tell you what I'd build. TOOLS Tracking: GA4 · Google Tag Manager · Meta Pixel · Meta Conversions API · Google Ads Data: BigQuery · Snowflake · PostgreSQL · AWS Lambda · Cloud Functions · dbt · Dagster · SQL · Python · Make BI: Power BI · Looker Studio · Tableau · Looker · Sigma · Metabase CRM: HubSpot · Salesforce · Pipedrive AI: Claude · ChatGPT · Gemini · Copilot Studio · Power Automate · Vellum