Technologies and methodologies I use to build data infrastructure for clients. --- > [!info] Consulting Stack > The default stack I deploy for mid-market engagements -- optimized for cost, speed, and maintainability. > > | Layer | Tools | > |-------|-------| > | **Extract & Load** | DLT (data load tool), Fivetran | > | **Staging** | Google Cloud Storage (GCS) | > | **Warehouse** | MotherDuck (DuckDB cloud), Snowflake | > | **Transformation** | dbt (Core & Cloud), SQL | > | **Analysis / BI** | Hex (notebooks + apps), custom React dashboards (Next.js), Evidence.dev; Tableau, Looker, Omni | > | **Semantic Layer** | dbt Semantic Layer / MetricFlow | > | **Orchestration** | Render cron jobs, dbt Cloud | > | **Version Control** | Git, GitHub | > [!info] Languages > > | Language | Context | > |----------|---------| > | **SQL** | Primary language - complex analytical queries, window functions, CTEs, Snowflake-specific optimization | > | **Python** | Data analysis, scripting, API integrations, automation | > | **YAML** | dbt model configuration, semantic layer definitions, CI/CD pipelines | > | **Shell / Bash** | Workflow automation, tooling scripts | > [!info] Analytics Engineering > - **dbt modeling patterns:** Staging → intermediate → marts; source freshness, schema tests, documentation > - **Semantic layer design:** MetricFlow metrics, dimensions, entities; governed metric definitions > - **Data testing:** dbt tests (schema, data, custom), CI/CD validation, data quality monitoring > - **Code review:** PR-based workflow for all model changes; documentation-as-code > [!info] Marketing Analytics > - **Funnel analysis:** Lead → MQL → SAL → SQO → Closed Won decomposition > - **Attribution modeling:** First-touch, multi-touch (time-decay), sourced vs. influenced frameworks > - **Pipeline forecasting:** Cohort-based conversion rates, time-lag modeling, scenario analysis > - **Campaign measurement:** Cost-per-MQL, pipeline influence, ROI by program type > [!info] Digital Advertising > - **Hands-on, 2012–2018:** LinkedIn Campaign Manager and Google Ads: audience building, bidding, creative testing, landing-page conversion > - **Since then, the data side:** attribution and campaign measurement, so spend is judged on pipeline and closed-won, not clicks > - **The loop:** run the campaign, land the results in the warehouse, read back what actually closed, reallocate > [!info] BI & Reporting > - **Dashboard design:** Stakeholder-first layout, progressive detail, consistent metric definitions > - **Platform migration:** Full lifecycle from audit → model redesign → build → training → adoption > - **Self-serve enablement:** Semantic layer integration, guided exploration, training programs > - **Code-based reporting:** Evidence.dev for version-controlled, DuckDB-native static dashboards > [!info] Workflow & Productivity > - **Claude Code:** Workflow automation, natural language orchestration, MCP server integration > - **Obsidian:** Knowledge management, structured task tracking (Bases plugin), Publish sites > - **Notion:** Project management, stakeholder-facing documentation > - **Jira:** Sprint planning, ticket tracking, accomplishment reporting --- ## Certifications & Learning - dbt Fundamentals (dbt Labs) - Snowflake ecosystem (hands-on daily use) - Continuous learner -- currently deepening expertise in LLM-assisted analytics workflows, lightweight EL pipelines (DLT), and cost-efficient warehouse patterns (MotherDuck/DuckDB)