dbt DAG Optimization Architect
Architect and optimize dbt DAGs for peak performance
What You Can Do
You analyze your dbt project's DAG structure to identify performance bottlenecks, circular dependencies, and architectural anti-patterns. Claude provides concrete materialization recommendations, dependency optimization strategies, and refactoring guidance to reduce build times and improve maintainability while preserving data lineage.
Features
identify which models constrain overall build performance
find and break problematic dependency loops that prevent incremental refreshes
decide between table, incremental, view, and ephemeral based on refresh cadence and downstream usage
pinpoint slow-running models and suggest optimization approaches (incremental loads, dbt+python, database-level optimization)
evaluate test distribution across DAG and identify untested critical nodes
auto-generate markdown or Mermaid diagrams showing data flows with business context
review custom macros for redundancy, performance, and compliance with dbt best practices
compare build metrics before and after changes to validate improvements
Example Output
Critical Path Analysis Report:
- Identified
fct_transactions(22 min) as critical path bottleneck - Recommended converting to incremental + partition by date (est. 3 min)
- Dependency map shows 8 downstream fact tables blocked by this single model
Materialization Recommendations:
model_name | current | recommended | reasoning
fct_daily_agg | table | incremental | 2GB daily volume, business needs fresh data hourly
dim_customer | table | snapshot | Slow-changing dimension, captures SCD Type 2 updates
stg_orders | view | ephemeral | Only consumed by 1 downstream model, ephemeral avoids materialization
Refactoring Workflow:
- Add
incremental+unique_keyto critical models - Extract shared logic into staging models (reduce redundancy)
- Move complex transformations to database views (push computation down)
- Add freshness SLAs to staging layer
What's Included
- SKILL.md: DAG audit workflows, optimization decision trees, and performance tuning playbooks
- DAG Analysis Template: structured checklist for parsing dependencies and identifying patterns
- Materialization Decision Matrix: table comparing refresh cadence, volume, and downstream usage patterns
- Performance Benchmarking Checklist: before/after metrics, profiling queries, and measurement framework
- Dependency Refactoring Guide: step-by-step workflow for breaking circular dependencies and reorganizing layers
Who It's For
- Analytics Engineering Leads — audit team dbt projects and enforce architectural standards
- dbt Architects — design scalable DAG structures for enterprise data platforms
- Data Platform Teams — optimize shared dbt projects used by multiple business units
- Senior Analytics Engineers — refactor complex projects to reduce technical debt
- Data Engineering Managers — benchmark build performance and identify team workflow improvements
Best For
- Auditing inherited or legacy dbt projects — quick assessment of architectural health and performance risks
- Redesigning DAG structure — breaking monolithic projects into modular layers (raw → staging → intermediate → mart)
- Optimizing slow builds — profiling and refactoring models that dominate CI/CD runtime
- Incremental model strategy — deciding which models should use incremental loads vs. full refreshes
- Documenting lineage and business context — generating lineage diagrams and data dictionaries
You might also like
Create compelling positioning statements, analyze competitive landscapes, and generate strategic brand recommendations backed by structured frameworks. You'll synthesize market data to identify differentiation opportunities, develop go-to-market strategies, and articulate your unique value proposition with confidence and clarity.
by lurenyang418
You generate production-ready LookML code, debug model validation errors, and optimize explore performance. This skill accelerates your development cycle by handling syntax generation, relationship definition, and complex calculations—letting you focus on business logic rather than language mechanics.
by lurenyang418
Analyze your current brand perception, competitive position, and market gaps to develop a comprehensive repositioning strategy. You'll receive a differentiated positioning statement, messaging architecture, competitive analysis, and a tactical 90-day implementation roadmap that your team can execute immediately.
by lurenyang418
This skill transforms raw competitive data—customer reviews, pricing structures, feature lists, and market reports—into structured, actionable intelligence reports. It creates competitive positioning matrices that reveal market gaps, pricing strategies, and feature differentiation at a glance. Get the strategic insights you need to inform product roadmaps, refine positioning, and make confident go-to-market decisions.
by lurenyang418
Transform raw qualitative data—interviews, surveys, focus groups, open-ended feedback—into structured themes and actionable insights. Claude systematically codes your data using established frameworks (Braun & Clarke, grounded theory, phenomenology), identifies patterns across respondents, and synthesizes findings into clear, evidence-based recommendations. You'll go from hundreds of transcript pages to organized themes with illustrative quotes in minutes.
by lurenyang418
You can create comprehensive, data-driven go-to-market strategies that align your enterprise product with market realities, competitive landscapes, and internal capabilities. This skill guides you through market analysis, positioning, timeline creation, pricing strategy, and launch readiness assessments—delivering actionable roadmaps that reduce launch risk and accelerate time-to-market for complex B2B initiatives.
by lurenyang418
You can extract, structure, and synthesize consumer insights from multiple research sources including interviews, surveys, and social listening data. This skill transforms raw qualitative and quantitative data into clear, actionable findings and detailed consumer personas that inform product, marketing, and business strategy decisions.
by lurenyang418
You upload interview transcripts, survey responses, or focus group data, and this skill systematically extracts themes, patterns, and relationships using proven thematic coding frameworks. It moves beyond raw data to deliver actionable strategic insights, complete with prioritization and visualization recommendations. The result is a structured, evidence-based foundation for decision-making.
by lurenyang418