Solutions Architect,
Commercial — dbt Labs

Mission

To empower data practitioners to create and disseminate organizational knowledge. dbt Labs pioneered analytics engineering — treating data transformation with the same rigor software engineers apply to application code.

Key Investors

Andreessen Horowitz, Sequoia Capital, Altimeter Capital, Coatue, Tiger Global, GIC, Databricks, Google Ventures, Snowflake.

Notable Customers

JetBlue, Nasdaq, AstraZeneca, Sky, Volvo, SafetyCulture, Siemens, Roche, Condé Nast, HubSpot.

Recent Milestones

• Jan 2025: Acquired SDF Labs — SQL comprehension for the dbt engine (foundation of Fusion)
• May 2025: Launched dbt Fusion Engine — Rust-based, up to 30x faster
• Sep 2025: Named to Forbes Cloud 100
• Oct 2025: Coalesce 2025 — dbt Agents, remote MCP Server GA, state-aware orchestration, Iceberg
• Oct 2025: Signed agreement to merge with Fivetran (all-stock, combined ~$600M ARR)
• Open-sourced MetricFlow under Apache 2.0 (Open Semantic Interchange initiative)

• Tristan Handy — Founder & CEO. 20+ years data practitioner. Analytics engineering newsletter & podcast. Post-merger becomes co-founder and President.
• Drew Banin — Co-founder & former CPO.
• Connor McArthur — Co-founder. Engineering background in data transformation.
• Sarah Riley — CFO. Previously Zoom (IPO → $4B revenue) and Okta.

Written before the company was founded. Deeply embedded in culture and interviews:

• We are a mission-driven company — Profits are a tool, not the mission.
• We contribute to the knowledge loop — Think in public, default to open source.
• We create value, not capture it — Optimize for the world's benefit first.
• Work done well is its own end — Craft and care matter intrinsically.
• Transparency always wins — Complete honesty, even when uncomfortable.
• We value diversity — Different mindsets create better outcomes.
• Users are our best advocates — Make users happy; they spread the word.
• We are humble — Excellence creates responsibility, not pride.
• We optimize for the long-term — Sustainable growth over short-term gains.
• We work hard and go home — Intense focus, then fully disconnect.
• We are human — Whole selves at work; identities extend beyond work.
• We believe in moving up the stack — Replace yourself through process; always grow.

dbt (data build tool) is a transformation framework that lets data practitioners write modular SQL (and Python) to transform raw data in a cloud warehouse into clean, tested, documented datasets.

Where dbt Fits

dbt is the "T" in ELT. It assumes raw data has been loaded (by Fivetran, Airbyte, Stitch). dbt transforms it. It does not extract or load data.

Two Engines

• dbt Core Engine — Open-source Python CLI. Renders Jinja, runs models. Apache 2.0.
• dbt Fusion Engine — Rust-based. Native SQL understanding. 30x faster. State-aware orchestration for cost savings.

Key Innovation: ref()

Reference models via {{ ref('model_name') }} instead of hardcoding table names. Creates a dependency graph (DAG) automatically, enables modularity, makes change propagation safe.

Models

SQL SELECT statements saved as .sql files. Each maps to a table or view. dbt handles DDL/DML and dependency ordering.

Materializations

Sources

Raw table declarations in YAML. Referenced via {{ source('schema', 'table') }}. Enable freshness checks and lineage.

Tests

• Generic — YAML: unique, not_null, accepted_values, relationships
• Singular — Custom SQL in tests/ returning failing rows

Other Concepts

• Seeds — CSV files loaded into warehouse
• Snapshots — Track historical changes (SCD Type 2)
• Macros — Reusable Jinja SQL snippets
• Packages — Via dbt Hub (dbt_utils, dbt_expectations)
• Exposures — Document downstream consumers
• Documentation — description: in YAML → dbt docs generate

my_dbt_project/
├── dbt_project.yml
├── models/
│   ├── staging/        # 1:1 with source, light cleaning
│   ├── intermediate/   # Business logic joins
│   └── marts/          # Wide, business-ready entities
├── macros/
├── tests/
├── seeds/
└── snapshots/

Source-conformed → Business-conformed: Staging mirrors sources; marts reflect business concepts.

-- ref() creates dependencies
SELECT order_id, customer_id,
  amount_cents / 100.0 AS amount_dollars
FROM {{ ref('stg_orders') }}

-- Conditional logic
{% if target.name == 'dev' %}
  WHERE created_at > dateadd('day', -3, current_date)
{% endif %}

{{ }} = expressions, {% %} = statements, {# #} = comments.

Next-gen Rust engine built on the SDF Labs acquisition (Jan 2025).

• SQL Comprehension — Validates syntax/logic before hitting the warehouse
• 30x faster parsing and compilation
• State-Aware Orchestration — Runs only changed models → 10%+ warehouse savings
• Column-Level Lineage — Traces columns through models
• LSP — Real-time error checking in VS Code and Cloud IDE
• Cost Dashboard — Track and optimize warehouse spend

Define metrics in YAML, query from any BI tool — consistent answers everywhere.

• Powered by MetricFlow (open-source Apache 2.0)
• GraphQL API, JDBC API, partner integrations
• Tableau, Looker, Power BI, Hex, Mode, Google Sheets

Why it matters: Creates a single source of truth for metrics across all tools. Huge enterprise selling point.

MCP server lets AI apps interact with dbt data assets. GA since Oct 2025.

• Local MCP — Full CLI, Fusion engine, optional platform APIs
• Remote MCP — Cloud-hosted; Semantic Layer, Discovery, SQL execution

dbt Agents (Beta)

• Developer — Code authoring, refactoring, validation
• Discovery — Find datasets, highlight trusted sources
• Observability — Monitor jobs, identify root causes
• Analyst — Answer questions about models and metrics

dbt Cloud ("dbt platform") is the managed SaaS — collaboration, orchestration, and governance on top of dbt Core/Fusion.

• Studio IDE — Browser-based. Dark mode, Canvas (drag-and-drop), Copilot (AI).
• VS Code Extension — Free local dev with Fusion. LSP-powered.
• Scheduling & Orchestration — Built-in scheduler, webhooks, API triggers.
• CI/CD — Auto-test on PR in temp schema. SQL linting.
• dbt Catalog — Lineage, metadata, global navigation.
• Semantic Layer — Consistent metrics to any BI tool.
• dbt Insights — Ad-hoc queries and data exploration.
• dbt Mesh — Cross-project references for federated architectures.
• Alerts — Email + Slack for job status.
• RBAC & SSO — Enterprise security, audit logging.
• Copilot — AI code gen, docs, tests, semantic models.

• Snowflake — Deepest integration. Partner of the Year 2025. Iceberg.
• Databricks — Lakeflow, Unity Catalog, Iceberg.
• BigQuery — DataFrames, Workload Identity Federation.
• Redshift — External OAuth via Okta/Entra.
• Also: Teradata, Fabric, Postgres, Trino, Salesforce Data Cloud, 30+ community adapters.

Must Be Fluent In

• CTEs, Window Functions (ROW_NUMBER, LAG, SUM() OVER)
• All JOIN types + EXISTS/NOT EXISTS
• Aggregations with CASE WHEN inside SUM/COUNT
• Date ops: DATEADD, DATEDIFF, DATE_TRUNC
• Type casting, string functions, COALESCE, NULLIF

Example: Incremental Model

{{ config(materialized='incremental', unique_key='event_id') }}

SELECT event_id, user_id, event_type, created_at
FROM {{ source('raw', 'events') }}
{% if is_incremental() %}
  WHERE created_at > (SELECT max(created_at) FROM {{ this }})
{% endif %}

• Separation of storage & compute — Scale independently. Industry standard.
• Columnar storage — Efficient for analytical aggregation queries.
• Warehouse: Structured, schema-on-write (Snowflake, BigQuery)
• Lake: Raw formats, schema-on-read (S3, GCS)
• Lakehouse: Open formats + ACID + SQL (Databricks, Snowflake w/Iceberg)
• Apache Iceberg — Open table format. dbt supports on Snowflake & Databricks.

• Star Schema — Fact + dimension tables. Classic BI.
• Wide/Denormalized — Common in modern warehouses.
• Data Vault — Hub/link/satellite. Enterprise-scale.
• One Big Table — Pre-joined wide tables for simple BI.

dbt supports all approaches. Staging → Intermediate → Marts is methodology-agnostic.

Direct

• SQLMesh — OSS alternative. Smaller community.
• Dataform — Google-owned. GCP-only.
• Coalesce — GUI-based. Snowflake-focused.

dbt's Moats

• 90K+ weekly teams — massive network effects
• Open-source foundation → bottom-up adoption
• "dbt" now a standard job requirement
• Fivetran merger → full ingestion-to-transformation

• Pre-sales (~70-80%) — Discovery, mapping dbt Cloud to customer needs
• Live demos — Tailored for technical and business impact
• Workshops — Deep dives on CI/CD, Semantic Layer, migration
• Partner with AEs — You own "how/why technically"
• Voice of customer — Feed insights to Product
• Content — Blogs, tutorials, event talks

Commercial segment: Mid-market, faster cycles, teams of 3–20. Often evaluating dbt Cloud vs. self-managing Core.

"Tell Me About Yourself"

Past → Present → Future → Why dbt Labs. Under 2 minutes.

"Why dbt Labs?"

Objection Handling Prep

• "We use Core for free" → TCO: infra, eng time, Fusion savings, Semantic Layer, security
• "Team lacks SQL skills" → Canvas, Copilot, free training, community
• "Locked into legacy ETL" → Coexistence, incremental migration
• "Too expensive" → ROI: Fusion cost savings, faster insights, less eng overhead

"Time You Failed"

Real example. Self-awareness + learning + application. Maps to humility and transparency values.

• Context (2 min) — "You face X, Y, Z..."
• Development (5 min) — IDE, model, ref(), tests
• Collaboration (3 min) — Git, PR CI, docs
• Deployment (3 min) — Scheduling, monitoring
• Lineage (2 min) — Catalog, DAG
• Value-add (3 min) — Semantic Layer or Fusion

Situation → Task → Action → Result (quantified). Prep 5–7 stories:

• Winning a customer with a demo/POC
• Collaborating with sales on a complex deal
• Handling a skeptical stakeholder
• Learning new tech quickly under pressure
• Improving a team process
• Navigating ambiguity

Team & Role

• "How is the Commercial SA team structured? AE:SA ratio?"
• "What does the first 90 days look like?"
• "How do you measure SA success?"
• "Balance of pre-sales vs internal work?"

Business

• "Most common customer profiles in Commercial?"
• "Biggest competitive challenge in deals?"
• "How will the Fivetran merger affect the SA role?"

Growth

• "Career paths for SAs?"
• "How does the Austin team use in-office days?"
• "What excites you most about the team right now?"