Dimensional modeling
Star schemas with conformed dimensions, so finance, sales and operations slice the same facts by the same customer, product and calendar.
Best forBI, self-service analytics and metrics people can trust.
Data architecture & engineering consulting
From data chaos to clarity.
We design the data architecture your business actually needs: models everyone trusts, warehouses built to last, and pipelines that don’t break every Monday. An architect-led team of data, cloud and DevOps engineers, for companies in Brazil, the US and beyond.
- Dimensional modeling
- Data Vault
- Data warehouse & lakehouse
- Snowflake & dbt
14+ yearsof the founder's work in data and technical leadership
Architect-led teamData, analytics, cloud and DevOps engineers, with a senior architect accountable end to end
Brazil · US · RemoteWorking sessions in English or Portuguese
SnowPro Advanced: Architectplus Databricks, dbt and AWS certifications
Sounds familiar?
Your company has plenty of data. What it’s missing is a design.
This is rarely a tooling problem. You can buy the best warehouse on the market and still get five answers to the same question, because nobody modeled what the question means.
Same word, five numbers
“Revenue” in finance isn’t “revenue” in sales. Every dashboard has its own version, and every meeting starts with an argument about which one is right.
Pipelines held together with duct tape
Scripts nobody owns, jobs that fail silently, and a weekly ritual of fixing last week’s loads before anyone notices.
History that quietly changes
A customer moves region or a product changes category, and last year’s reports change with it. Nothing is reproducible, nothing is auditable.
A platform that grew by accident
A warehouse here, a lake there, a few tools on top. The cloud bill keeps growing and nobody can draw the whole picture on one page.
Tools don’t fix this. Architecture does. And no, AI won’t fix it either: AI can’t fix chaos, it needs structure.
Our approach
Let’s be honest: most data projects start with a tool and end with a mess. We start at the other end.
First, how your business works and what it needs to know. Then the model that makes each answer true everywhere: one customer, one order, one revenue, across every system and every report. Only then the tools and the pipelines, built together with your team so the knowledge stays in-house.
What you have
- crm.contacts_v2_final
- erp.TB_CLI_001
- billing.customers (3 sources)
- revenue_FINAL(2).xlsx
- “Revenue” dashboard × 5 versions
What you get
- dim_customer · one customer, every system
- fact_sales · grain: order line
- hub_customer + sat_* · full history
- net_revenue · one definition
- lineage & tests on every table
The model is the product
Real data warehousing, modeled across your whole company.
We use proven modeling techniques, picked for the problem rather than for fashion, and often combined in layers.
Data Vault
Hubs, links and satellites that integrate many sources, absorb change without rework and keep the complete, auditable history of your business.
Best forEnterprise integration layers, many changing sources, audit and compliance.
Lakehouse architecture
Layered bronze, silver and gold zones on open table formats such as Apache Iceberg, mixing warehouse discipline with lake-scale storage.
Best forLarge or semi-structured data, ML and AI workloads, avoiding lock-in.
What we do
From architecture to the last mile of your data.
Six service areas, one way of working: design first, build it right, leave your team stronger.
Data Architecture & Strategy
A target architecture and a roadmap you can actually execute, from warehouse and lakehouse design to integration patterns and technical governance.
Learn more →Data Modeling & Governance
Dimensional models, Data Vault and enterprise models that make every team count the same customer, the same order and the same revenue.
Learn more →Data Engineering
dbt-centered transformation, ELT and ETL pipelines, batch and streaming, orchestration and data quality checks built to run quietly every day.
Learn more →Cloud Data Platforms
Snowflake specialists, fluent in Databricks, BigQuery, Redshift, Dremio and open-source stacks on AWS, Azure or Google Cloud, with performance and cost under control.
Learn more →Technical Advisory & Mentoring
Architecture reviews, design decisions and hands-on mentoring for data teams that want to level up how they build.
Learn more →AI & Data Integration
AI-ready data foundations: governed, modeled data that assistants, retrieval systems and agents can use without guessing.
Learn more →
Your options
There are three ways to get a solid data foundation.
Same goal, very different costs, timelines and odds of getting the architecture right the first time.
| Criteria | Build a full in-house team | Hire a large consultancy | Vitridata |
|---|---|---|---|
| Who designs the architecture | Whoever you manage to hire, once you hire them | A senior in the pitch, often a junior team on delivery | A senior data architect, from the first call to delivery, backed by our engineering team |
| Time to start | Months of recruiting and onboarding | Weeks of proposals and staffing | Usually within a couple of weeks |
| Modeling depth | Depends on one or two hires | Often template-driven | Dimensional, Data Vault and lakehouse, chosen per problem |
| Tool choices | Whatever the team already knows | Often tied to partner agreements | Whatever fits your data: Snowflake, Databricks, BigQuery, Redshift, Dremio or open source, on AWS, Azure or Google Cloud. We go deepest on Snowflake and dbt, and we sell no licenses, so our advice has no hidden agenda. |
| Cost profile | Fixed payroll, whether or not the work is there | High rates plus overhead | Senior expertise sized to the work: project, retainer or contract hours |
| What stays with you | Knowledge that leaves when people do | Deliverables, sometimes a dependency | Documented architecture, trained team, no lock-in |
Selected results
What good architecture looks like in numbers.
Results from past engagements led by our founder. Client names are kept confidential.
70%Faster processingMigration from AWS to Databricks
45%Lower platform costOptimization of AWS data environments
80%Faster data loadsPipeline execution-plan analysis and tuning
100%Lineage coverageAutomated metadata layer
30 daysMaster data managementComplete MDM implementation
90 daysIntegrated data warehouseFrom disconnected sources to one warehouse
Certifications
Certified where it counts.
Including Snowflake’s most senior architecture credential. We work across platforms and have a clear preference for the Snowflake ecosystem and dbt, where our expertise runs deepest.
SnowflakeSnowPro Advanced: Architect
SnowflakeSnowPro Core
DatabricksDatabricks Certified Data Engineer Associate
dbt Labsdbt Analytics Engineering Certification
Amazon Web ServicesAWS Certified Cloud Practitioner
EXINInformation Security Foundation based on ISO/IEC 27001 (ISFS)
Individual professional certifications held by our team. They are not vendor partnerships.
How we work
A clear path, flexible enough for the real world.
What we need from you: a few conversations with the people who know the business, and access to the systems. We take it from there.
Discover & assess
We map your sources, models, pipelines, costs and the questions the business can't answer yet.
Design the architecture
A target architecture and data model, with the trade-offs written down and a realistic roadmap.
Implement & integrate
We build the models and pipelines with you, in small increments that deliver value early.
Validate & optimize
Data quality tests, reconciliation with the business, performance and cost tuning.
Transfer & support
Documentation, training and mentoring so your team owns it. Ongoing support if you want it.
Do you work with companies in the United States?
Yes. We work remotely with companies in the US and elsewhere, in English. Brasília time (UTC−3) overlaps most of the US business day, so live working sessions are easy to schedule. We can engage on a project basis or as an independent contractor embedded in your team.
Are you tied to a specific vendor or tool?
No. We are cloud- and tool-agnostic and have no reseller agreements. We have delivered on Snowflake, Databricks, BigQuery, Redshift, Dremio and fully open-source stacks, on AWS, Azure and Google Cloud. We do have a preference: the Snowflake ecosystem with dbt, where we hold Snowflake’s advanced architect certification. But the recommendation always follows your data, your team and your budget.
Dimensional model or Data Vault: which one do we need?
Often both, in different layers. Data Vault is great for integrating many changing sources with full history; dimensional models are what analysts and BI tools should read. We decide based on your sources, your audit needs and who consumes the data, and we explain the trade-offs in plain language.
We already have a data platform. Can you help without starting over?
Most of our work starts exactly there. We assess what exists, keep what works, and fix the model, the pipelines or the costs where it matters most. Rewrites happen only when they clearly pay off.
Who will actually do the work?
An architect-led team. A senior data architect owns the design and stays with you through delivery, supported by our data engineers, analytics engineers and cloud/DevOps engineers as the project needs. You always know who is on the team and why, and you are never handed off to a junior bench after the sales call.
How do engagements and pricing work?
It depends on the shape of the work: a fixed-scope project, a monthly advisory retainer, or hourly contract work embedded in your team. After a first conversation we send a written proposal with scope, timeline and price.
Will our team be able to run it after you leave?
Yes, that is the plan from day one. Knowledge transfer is part of every engagement: your engineers pair with ours, we run hands-on workshops on the models and pipelines we deliver, and we document architecture decisions, data dictionaries and runbooks. Ownership moves to your team step by step, and we stay available for support if you want us.
Do you also work with AI?
Yes, from the data side. AI can't fix chaos: it needs structure. We build the governed, well-modeled data that AI assistants, retrieval systems and agents depend on, and we design how they access it safely.
Get started
Let's build your data foundation.
Tell us where your data hurts today. In the first conversation you talk to the architect who would do the work, and you leave with a clearer picture, whether or not we work together.