01
Ingest
Streaming, CDC, batch loaders. Contracts at the edge so bad upstream data doesn't poison the model layer.
Fivetran · Airbyte · Kafka · custom
Your data teammate · from ingest to dataviz
Ship trusted business metrics in weeks, not quarters.
TwiceData delivers production-ready dbt model packs, change-safe lineage controls, and a governed semantic layer so VP Data teams stop rewriting the same metric logic every quarter.
Why TwiceData
The same metrics, rebuilt at every company.
Every finance team re-derives ARR and re-fights gross-vs-net. Every Medicare Advantage plan re-implements CMS-HCC RAF. Every marketplace re-argues GMV and principal-vs-agent. Every EU company is about to re-learn ESRS and EHDS the hard way. It's identical, expensive, error-prone work — written from scratch, four times, in four companies.
We made the defensible answer once. TwiceData packs are drop-in, governed, source-agnostic dbt models. Install one and get best-practice, compliance-aware marts in days, not quarters — eight domains, multi-engine, audit-ready.
Across your data pipeline
Most consulting shops audit one layer and walk away. We sit across all five.
Modern data teams touch the same five layers every week — ingest, store, model, semantic, dataviz. TwiceData engages at any layer, owns the seams between them, and stays past handoff. The result is one teammate, one accountability, one working pipeline.
02
Store
Warehouse choice, partition strategy, materialization plan, cost controls. The bill where most data teams bleed.
Snowflake · BigQuery · Redshift · Databricks · DuckDB
03
Model
Governed, tested transformations and reusable domains — dbt where it's the right call (it usually is), or whatever framework your team already runs. The metric definitions that have to survive across teams.
dbt · SQLMesh · Dataform · Spark · SQL + Python
04
Semantic
Certified metrics, semantic layer, RBAC. The single source of truth that finance, product, and the board can all sign.
Cube · LookML · dbt Semantic Layer
05
Dataviz
Executive dashboards, ad-hoc exploration, reverse-ETL back to operational tools. The output your team actually reads.
Looker · Mode · Hex · Tableau · Metabase
00 · Foundation
Infrastructure & orchestration
The platform the five layers run on. We stand it up as code — provisioned, scheduled, shipped through CI/CD — so the whole pipeline is reproducible from day one, not click-ops nobody can rebuild.
Terraform · Pulumi · Airflow · Dagster · Prefect · GitHub Actions / CI
Engage us across all five layers, or on just one or two — you choose the scope. Most engagements start at the layer where the pipeline is broken and grow from there; some stay focused on the single layer that needed help. The shape is yours.
Prebuilt domain packs
Your domain, modeled on day one.
Each pack is three layers: a governed model pack (dbt-first, or your framework), a domain knowledge graph, and an MCP server so your AI agents query the governed data directly. The models are table stakes — the graph and the MCP are what your competitors don't ship. All eight domains — Finance, Sales & Marketing, Healthcare, Legal, Public sector, SaaS, and Marketplaces, plus an Enterprise governance meta-layer — are built and shipping, with EU packs (CSRD/ESRS and EHDS) and a shared temporal toolkit underneath; portable to SQLMesh, Dataform, or pure SQL.
Finance
A financial-metrics knowledge graph and a finance MCP — certified ARR/MRR, cohorts, and revenue recognition exposed to AI agents as governed tools.
Foundation: ARR/revenue dbt model pack
Finance domain →Medical
A risk-based clinical knowledge graph and a medical MCP server — your AI agents query governed, HIPAA-aware patients, claims, and HCC risk directly, not raw SQL.
Foundation: risk-adjusted dbt model pack
Healthcare domain →Legal
A matter & contract knowledge graph with legal-document intelligence, served over MCP — clauses, obligations, and parties your AI agents can traverse.
Foundation: billing & realization dbt model pack
Legal domain →Sales & Marketing
A revenue & attribution knowledge graph, served over MCP — pipeline, multi-touch attribution, and CAC/LTV your AI agents can query.
Foundation: pipeline & attribution dbt model pack
Sales & Marketing domain →Public sector
A government-finance & program knowledge graph and a public-sector MCP — TAS-keyed budget execution, sub-award lineage, and CIPSEA-aware suppression your AI agents query under FISMA-aware controls.
Foundation: budget-execution dbt model pack
Public sector domain →SaaS
A product-led-growth knowledge graph and a SaaS MCP — activation, engagement (DAU/WAU/MAU), feature adoption, and PQL scoring exposed to AI agents as governed tools.
Foundation: product-engagement dbt model pack
SaaS domain →Marketplaces
A two-sided-economics knowledge graph and a marketplace MCP — GMV, take-rate, liquidity, and trust & safety your AI agents query without re-arguing principal-vs-agent.
Foundation: GMV & liquidity dbt model pack
Marketplaces domain →Enterprise
A governance meta-layer over every pack — lineage, cost/FinOps, contract compliance, and a unified semantic layer, served over an MCP so agents can ask the platform anything.
Foundation: dbt-artifact + warehouse-telemetry pack
Enterprise domain →Pillar 01
Pre-built dbt model packs
Start with curated SaaS subject areas for ARR, retention, product usage, and revenue operations. Each pack includes tests, documentation, and versioned assumptions.
Pillar 02
Lineage-aware change management
Every model update is impact-scored before merge. Downstream dashboards and semantic entities are flagged automatically, so breakage is caught before it reaches executives.
Pillar 03
Governed semantic layer
Define certified metrics once and expose them to BI, reverse ETL, and notebooks with role-based access, freshness policies, and audit history built in.
Before: fragmented stack
Local definitions + manual handoffs
Warehouse raw tables Source
Team-specific dbt forks Divergent
BI semantic logic per dashboard Duplicated
Spreadsheet reconciliations Manual
Metric disputes in exec review Weekly
After: managed model delivery
Single governed pipeline
Curated model packs Versioned
Lineage impact checks Pre-merge
Certified semantic entities Reusable
Access + policy controls Governed
Consistent board-level KPIs Trusted
Need project-specific help without adding permanent headcount?
For small and growing data-driven teams, we can be your full data team, work alongside your existing team, or deliver a scoped project your team can run and ingest without a long hiring cycle.
Service 01
Stack Diagnostic
Cross-system audit of warehouse, orchestration, BI, and activation layers to isolate cost leaks, model sprawl, and metric drift.
- Lineage + model complexity scorecard
- Warehouse and cloud spend-to-value map
- 30-day remediation plan
Service 02
Model Efficiency Upgrade
Refactor slow or duplicated transformations into governed reusable domains with stronger test coverage and cleaner ownership.
- Query performance optimization
- Model consolidation and naming cleanup
- dbt and semantic-layer hardening
Service 03
Embedded Delivery Sprint
Boots-on-keyboard support from senior data and cloud engineering for one strategic project, delivered under SOW.
- 2-6 week contract engagements
- Weekly deliverables and stakeholder demos
- Handoff docs + enablement session
Product 01 · Flagship
Quarter Stack
Turnkey 12-week build of your full data model + analytics + visualization layer. Custom or Express track. Handoff or subscription.
- Full model · analytics · viz stack
- $27K–$41K build · Custom or Express track
- Two tech tracks: Custom or Express
Technology ecosystem
We orchestrate across your existing stack and deliver project milestones with your team, not around it.
Data platforms
Cloud + enterprise
Delivery model
Contract-first engagement We deliver scoped outcomes with defined milestones, whether you need a full external data team, augmentation for your current team, or a project handoff your team can operate immediately.
Beyond the prebuilt packs: custom by design.
The eight packs cover the metrics every company in a domain rebuilds. When your data doesn't fit a pack — a novel domain, a bespoke metric, a regulatory edge — we design it with you, with the same governed, source-agnostic, audit-ready rigor, built to your spec.
Novel domains
Your vertical, modeled
No pack for your industry yet? We build the source contract, governed marts, knowledge graph, and MCP from scratch — the same architecture the prebuilt packs are built on.
Bespoke metrics
The number only you compute
A proprietary metric, an unusual revenue model, a custom risk score — defined once, governed, and certified so it survives across teams and audits.
Regulatory edge
Compliance to your jurisdiction
Region- or regulator-specific requirements — EU data residency, sector rules, internal audit — modeled in, with redaction, lineage, and audit-readiness built into the pipeline.
Start here · the build
Most engagements start with a build.
A fixed-scope engagement — a stack audit, an embedded sprint, or a full quarter build, scoped to your budget. We build your governed data layer and hand it off, documented and tested. See engagements & pricing →
Optional · after the build
Keep it running
Once your build is live, an optional subscription keeps your model packs and governance current.
Starter
$1,500/month
- 2 model packs
- 1 production workspace
- Email support
Team
$2,900/month
- 6 model packs + semantic governance
- Lineage-aware change approvals
- Slack + office hours support
Enterprise
Customannual
- Unlimited domains and entities
- SSO, RBAC, and policy controls
- Dedicated success architect
AI consulting and engineering for teams that want to move, but need a safe starting point
For companies that are AI-curious but hesitant to execute, we scope practical integrations tied to real workflows, risk controls, and measurable outcomes. We can lead end-to-end, co-deliver with your existing team, or ship an AI project your team can use and ingest with confidence.
- AI readiness assessment across data quality, access controls, and system architecture.
- Use-case prioritization for internal copilots, analytics assistants, and workflow automation.
- Pilot build and evaluation framework with governance guardrails and rollout checkpoints.
First-hour consultation: free We offer a no-cost first-hour AI integration consult to map fit, constraints, and a realistic path to production.
Replace metric drift with a governed model system.
See your stack mapped to managed model packs and a rollout plan in a 30-minute architecture session.