Restaurant
- Sales
- Inventory
- Purchasing
- Bookings
- Customers
- Promotions
Your company already produces the data it needs.
The problem is that it's scattered.
- Systems
- Spreadsheets
- Emails
- Documents
- Conversations
- Purchases
- Sales
- Contracts
- Equipment
- People
GSA organizes these signals into a structure that helps people understand, ask, decide and act.
01Approach
The problem isn't a lack of data.
Every operation talks all day long. Usually in different places, under different names, kept by different people.
Construction
- Budget
- Planning
- Purchasing
- Measurements
- Productivity
- Execution
Environment
- Permit conditions
- Lab analyses
- Documents
- Deadlines
- Correspondence
Legal
- Cases
- Rulings
- Documents
- Case law
- Deadlines
- History
Engineering consultancy
- Contracts
- Proposals
- Projects
- Scopes
- Hours
- Technical knowledge
The data is there. What's missing is bringing it together in a way someone can actually use.
02
Everyone wants the answer data can give.
Few people want to stop and feed the data.
That's why a large part of our work happens before any analysis: making capture easier. Fewer spreadsheets. Less retyping. Fewer people copying information from one place to another.
- Voice
- Documents
- Emails
- Systems
- Images
- Sensors
- Operations
Whenever possible, data should be born from the work itself. Not from extra work done just to produce data.
03
Talking to the business.
Once data and knowledge are organized, asking becomes simple again.
You ask in your own words. The answer comes back in your own words.
- How much did I really make last month?
- Why did my margin drop?
- Which products aren't moving?
- Which past contracts look like this opportunity?
- Which permit conditions expire in the next 60 days?
- Which sites are below planned productivity?
- What has changed in my customers' behavior?
04
From Data Analyst 1.0 to Data Professional 2.0.
The role didn't end. It moved.
Data Analyst 1.0
Much of the time went into just getting to the information:
- preparing data
- fixing spreadsheets
- writing queries
- updating dashboards
- building charts
- repeating operational tasks
Data Professional 2.0
Technology takes over part of that work. The time left goes to what always mattered:
- understanding the business
- asking good questions
- interpreting evidence
- finding opportunities
- supporting decisions
This isn't about replacing people. It's about giving them back the work that requires thinking.
05
We don't start by asking which dashboard you want.
We start by asking which problem you need to understand or solve.
- 01What happens in the operation?
- 02What signals exist?
- 03Which data is already produced?
- 04Which still needs to be captured?
- 05Who can access each piece of information?
- 06Which decisions need to be made?
- 07What can be automated?
- 08What should remain a human decision?
Technology comes after the problem. Never before.
06
What we build.
This isn't a catalog. These are capabilities an organization comes to have.
Management Intelligence
Revenue, costs, margins, operations, customers, productivity and targets. With the option to ask in plain language.
Service and Relationships
Systems that answer, guide, book and win back opportunities. Waiting for the customer or reaching out.
Corporate Knowledge
Documents, contracts, proposals, standards and records that are scattered today, turned into information you can find.
Monitoring and Forecasting
Patterns, anomalies, trends, risks and opportunities noticed before they become surprises.
Smart Data Capture
Collection built into the process itself, so nobody has to stop working to record the work.
Sometimes the result is a dashboard. Sometimes it's an automation, a statistical model, a system, a report or a WhatsApp agent. The format is a consequence. What matters is the capacity to understand, decide and act.
07Technology
What's under the hood.
For those who like to know what things are called.
You don't need to understand any of this to work with GSA. But if you want to, here's the vocabulary, in plain English.
We don't publish our project architectures. Each client gets their own, designed for their problem.
-
01
Databases
The company's organized memory.
Sales, inventory, contracts and hours stop living in loose spreadsheets and get a fixed address, a history and rules. Without that, any AI is just guessing.
-
02
APIs Application Programming Interfaces
Systems talking to each other.
ERP, CRM, spreadsheets, WhatsApp and databases exchange information on their own, with nobody copying and pasting. It's how data comes in and how decisions go back into operations.
-
03
Embeddings
Search by meaning, not by word.
Text becomes coordinates of meaning. So a search for “site delay” also finds “blown schedule” and “late delivery”.
-
04
RAG Retrieval-Augmented Generation
The AI checks before it answers.
Instead of answering from memory, the AI looks through your documents and data, reads the right passage and answers showing where it came from. Less invention, more evidence.
-
05
Contextualized AI
An AI that knows your business.
It knows what an “order” means to you, who can see what and which numbers count. When needed, it runs inside your own company, and the data never leaves the building.
-
06
MCP Model Context Protocol
AI's universal plug.
An open standard that connects AI to the company's tools and data in a controlled way. It defines what the AI can look up and what it can do. And nothing beyond that.
-
07
AI agents
AI that acts, within limits.
They answer, schedule, check, alert and prepare. Always with clear rules, a record of what they did and a person nearby for the decisions that matter.
- Human decision
The acronyms change every year. The rule doesn't: problem first, technology second.
08Applications
Where we apply our approach.
The reasoning is the same. The signals change from sector to sector.
Restaurants, entertainment and sports courts
Bookings, inventory, purchasing, pricing, promotions, relationships, occupancy and management.
Medical and dental clinics
Service, scheduling, relationships, operations and institutional knowledge, with the care sensitive data requires.
Engineering and construction
Budget vs. actual, productivity, measurement, purchasing, planning and field records.
Environment and mining
Permit conditions, deadlines, documents, lab analyses, limits, reports and regulatory knowledge.
Engineering consultancy
Contracts, proposals, scopes, records and recovering the company's technical knowledge.
Legal
Document organization, research, case law and quantitative analysis of case histories.
Retail and supermarkets
Sales, inventory, margin, stockouts, promotions, shopping behavior and demand forecasting.
Facilities and outsourced services
Contracts, teams, shifts, productivity, training, tenders and operations.
Startups
A data-driven operation from day one, not patched on later.
Sectors where we apply the method. It doesn't mean there's a ready-made product for each one.
09Clients
Clients served by GSA
Different sectors, the same starting point: a real problem and data scattered across the operation.
10
The principle is the same. The scale isn't.
GSA fits the dose to the size of the problem. No more, no less.
11GSA Lab
Where we explore what data will be able to do tomorrow.
Applied research and experiments. Some become method. Others become lessons.
- 01Computer vision
- 02Data from cameras and sensors
- 03Territorial intelligence
- 04New interfaces between people and AI
- 05AI applied to engineering
- 06Modeling and simulation
- 07Automatic capture
- 08Specialized agents
- 09Local models
- 10New ways to talk to data
What lives in the Lab is an experiment, not a product.
12Education
Learning to think Data-Driven.
Tools can be learned from a tutorial. The hard part is the reasoning: which question to ask, which data to look at, when to trust the number and when to doubt it.
Courses, articles and videos in preparation.
- Data-driven restaurants
- Data-driven construction
- Evidence-based management
- AI applied to business
- Data-Driven startups
- Data Professional 2.0
- Smart data capture
- Case studies and real experiences
13Observatory
GSA Data & AI Observatory
Public data about the market, produced by us.
If we argue for evidence-based decisions, we need to produce evidence too.
First project
The Data & AI job market in Brazil.
Salaries, median and quartiles
Ranges by level: Junior, Mid-level and Senior
Distribution by state
Roles and emerging job titles
Skills requested in job postings
Technologies mentioned
Change over time
Under construction. No figures will be published until there's a real, verifiable dataset.
14Library
Open knowledge from Data-Driven Culture and GSA.
From scattered information to organizational intelligence
In preparation10 questions to find out whether your company really uses the data it already has
In preparationWhen the materials are available, we'll only ask for what we need to send them and stay in touch. We collect data because there's a purpose, not because we can.
15About
Engineering taught me to observe real systems.
Data taught me to measure them.
AI expanded our ability to interpret them and act on them.
Data & AI Strategy Consultant
Consultor e Estrategista de Dados e IA
A civil engineer who moved into data without leaving the job site behind. GSA was born where operations, business, data and AI meet.
Every organization produces signals.Some disappear.Others become data.Few ever turn into better decisions.
That's the path GSA works on.
16Contact
Shall we talk about a problem?
You don't need to arrive with a ready-made solution. You can start with the problem, with the data you have, or just with a question.