JTI
During my Business Intelligence internship, I worked with nationwide sales data, Excel dashboards and analyses for senior management and commercial teams.
Explore my work with dataOperations, processes
and data analysis.
I am a production engineer with experience in commercial analysis, business management and automotive diagnostics. My work has involved analyzing data, organizing services and working with customers, teams and suppliers.
I started in Business Intelligence at JTI. I then founded Autostatt and managed the workshop. At Race Performance, I continued my technical work in automotive diagnostics.
During my Business Intelligence internship, I worked with nationwide sales data, Excel dashboards and analyses for senior management and commercial teams.
Explore my work with dataI founded and ran the business during this period. I handled customers, estimates, the technical team, purchasing and service delivery.
Explore my operations experienceI worked in automotive diagnostics, from gathering symptoms and conducting tests to validating repairs and communicating the findings.
View work from this periodMackenzie Presbyterian University
Native Portuguese · Fluent English · One year of high school in the United States
Engine Assembly and Metrology
Engine Performance
Two semester-long modules completed over one year, with 7 hours of classes per week.
At JTI, I worked across the sales reporting process, from data quality to updating performance indicators and analyzing results.
I processed nationwide sales data, identifying duplicates and missing information.
I updated Excel dashboards daily, with views by salesperson, supervisor, branch and Brazil overall.
I analyzed performance, sales incentives, product mix and new products to support senior management and commercial teams.
I founded and ran Autostatt from 2018 to 2023. Alongside the technical work, I managed day-to-day operations: customers, staff, purchasing, suppliers and deadlines. I am no longer the owner.
Founder and operations lead
during this period
I spoke with customers to understand their needs and prepared service estimates.
Customer · Scope · EstimateI scheduled jobs and coordinated the technical team, setting priorities and deadlines.
Planning · Team · DeadlinesI managed purchasing, supplier relationships and stock. I assessed the cost, availability and compatibility of parts, as well as their impact on delivery times.
Purchasing · Suppliers · InventoryI tracked progress and deadlines through to handover. I used spreadsheets to keep the operation organized.
Execution · Tracking · DeliveryWhile running Autostatt, I hosted video crews to show the workshop in action and explain our services. Here are two of those recordings.
In this appearance on Canal da Peça, I explain the differences between 3D and conventional wheel alignment. We filmed the video at Autostatt in 2019.
Watch on YouTubeI hosted Auto Direto at Autostatt to document an inspection of a Fusion EcoBoost’s intake valves and the walnut-blasting carbon cleaning process.
Watch on YouTubeThe original videos are in Portuguese. If playback is unavailable here, use the YouTube links.
I welcomed the Oficina Brasil team to Autostatt and shared my assessment of the Renault Sandero RS. My comments appeared in the article published on 12 March 2023.
This was a vehicle assessment for the article; I did not perform a repair on that car.
Photos and videos from the business I ran, including the workshop, repairs and services.
Photos: Oficina Brasil. Article by Antônio Gaspar de Oliveira, published on 12 March 2023.
I have collected customer comments about diagnostics, service and confidence in Autostatt’s work.
“Em dois dias meu carro estava pronto e funcionando novamente, com as anomalias totalmente sanadas.”English translation“Within two days my car was ready and running again, with all the issues fully resolved.”
Read the original review on Google MapsDate shown by Google: 4 years ago.
“Após uma longa conversa me senti seguro e decidi aprovar o serviço.”English translation“After a long conversation, I felt confident and decided to approve the work.”
Read the original review on Google MapsDate shown by Google: 8 years ago.
“Não ficaram empurrando serviços desnecessários e o valor cobrado foi justo”English translation“They did not push unnecessary services, and the price charged was fair.”
Read the original review on Google MapsDate shown by Google: 5 years ago.
“Tudo isso nos deixa tranquilos em realizar todo tipo de manutenção e reparo na Autostatt.”English translation“All of this gives us peace of mind about having any type of maintenance or repair done at Autostatt.”
Read the original review on Google MapsDate shown by Google: edited 5 years ago.
Cropped original screenshots viewed on 1 October 2026, preserving the stars, relative dates and original Portuguese text. Reviewer names and avatars are outside the crops. Links lead to the public reviews. English translations of the excerpts are provided below the original quotations.
“Gabriel obrigado pelo excelente serviço prestado na manutenção do meu Jetta TSI!”English translation“Gabriel, thank you for the excellent work servicing my Jetta TSI!”
Open the YouTube commentComment on the Canal da Peça video. The name and avatar are outside the screenshot.
“Gostei do serviço e preço ok”English translation“I liked the service and the price was okay.”
Open the Facebook conversationFeedback in an automotive community. The name and avatar are outside the screenshot.
Comments posted on YouTube and Facebook. Select the images to enlarge them. Original Portuguese excerpts are followed by labeled English translations.
These examples show how I approached diagnostics and what I did on each job. The photographs and videos come from my professional archive.
I continued using @autostatt_ to share my work after leaving the business. The cases below are from that period; the intake valve cleaning was carried out at Race Performance.
I investigated a Golf GTI that produced smoke only at idle or under no load.
I began work on this Audi Q5’s EA888 engine because of a leak.
I carried out intake valve carbon cleaning on a direct-injection engine at Race Performance.
Dates refer to posts on my @autostatt_ profile. Select a photograph to enlarge it.
I carried out the performance work on this Jetta GLI at Race Performance. I shared it on my profile in November 2024, with photographs and a video of the car on the dynamometer.
In the recording, I explain the build beside the car. In the post, I described the project as Stage 3 for use with regular gasoline.
Watch my explanation on InstagramVideo and photographs from the post dated 7 November 2024. The Portuguese video opens on Instagram, which may require login.
I am studying how AI can connect hands-on technical knowledge with workshop decisions and routines. DiagFlow is the broader vision; in AllMotor, I am developing and testing a focused part of that idea.
Mechanics generate essential information. Repeating explanations, filling in forms and responding to interruptions competes with work on the vehicle. I want to turn those observations into useful information and actions that can be tracked.
My idea is to use the work order context to organize observations, identify gaps, suggest hypotheses and next tests, and prepare requests for other teams. The mechanic chooses when to seek guidance and when simply to record what they have found.
I started with DiagFlow, but the project’s complexity led me to revisit the scope. In AllMotor, I began working in stages, using the existing customer, vehicle, estimate, work order and history records.
In the local test version, I am developing editable proposals, structured observations, summaries and explanations. Proposals go through human review before adoption, with authorship recorded. Full integration across teams remains part of the longer-term plan.
These five screens from the local version show how I am organizing technical observations, proposal review and service history.
Local demo · Fictional data · Simulated AI · Learning project in progress
I keep the reported issue, vehicle and planned work together in the work order context.
This screenshot shows the stage before a proposal, while the diagnosis is still open.
I separate observations, hypotheses and suggested tests before applying a proposal to the form.
The response is simulated. Source references point to the observations under review; suggested tests require authorization.
I retain the link between the recorded finding and the reviewed proposal in the history.
An automated test went through the adoption process. The record shows authorship and the linked proposal, without confirming a diagnosis.
I bring together the customer, vehicle and work order to recover the context of the service.
A button to summarize service records is available; this screen does not show a generated summary.
I connect the vehicle to its service records and the option to summarize its history.
This screenshot shows the linked records and the summary entry point, without a generated AI response.
These are genuine screenshots of the local interface, with fictional data in an isolated environment. The proposal shown uses a fixed test response, without calling a live model. An automated test exercised review and adoption. I am still evaluating how clear and useful this workflow is for everyday workshop use. The original Portuguese interface is preserved in the screenshots.
I want information to follow the work across the workshop, purchasing, customer service and management, with clear ownership and visible responses. AI can help prepare requests and customer explanations; technical and commercial decisions remain with the people responsible.
Record observations and evidence in the context of the vehicle and work order
Review interpretations and define the next tests or actions
Turn confirmed information into requests for the teams involved
Track each request, its owner and the response back to the mechanic
I want to keep hypotheses, evidence and conclusions clearly distinct. The professional can change the investigation sequence, add their own tests, revise interpretations and leave a case unresolved.
I am learning to turn this vision into verifiable workflows, preserve context when AI fails and distinguish a useful suggestion from a confirmed decision. Results and decisions need traceable sources, authorship and history.
This is an ongoing learning project. My next step is to review complete scenarios and assess the clarity and usefulness of the proposals.
Gabriel Cappello Machado
I am seeking opportunities in operations, processes and data analysis. Based in São Paulo, I bring experience in Business Intelligence, business management and technical work.
This collection brings together my professional archive, Autostatt posts, and appearances in videos and articles. Each piece of work links to its original publication.
The photographs of the Autostatt visit are from Oficina Brasil’s article by Antônio Gaspar de Oliveira, published on 12 March 2023. The featured videos are by Canal da Peça and Auto Direto. The Subaru video was produced by Hericles Oliveira.
Dates refer to the original publication dates. Customer names and avatars were omitted from comment screenshots; license plates were obscured in the exterior and visit photographs. Some external links may require login. Photographs and videos remain the property of their respective creators. Original media and source publications are in Portuguese.