// PROCESSES → DATA → DECISIONS

I turn processes
into systems that
remove noise.

Data & Automation Analyst with 6+ years across the automotive ecosystem. I investigate the process, find the bottleneck and build the solution — from operational automation to pipelines, integrations and systems.

EXPERIENCE / ECOSYSTEM
Stellantis General Motors MSX International Conduent Infosys
// REAL IMPACT2026
01 DAILY NPS
~13 min~20 s

human approval remains at the decision point

02ORACLE VALIDATION~3 min → ~34 s
03POSTCODES PROCESSED+1M
04AUTOMATED BASES4
CLARITY · EXECUTION · TRACEABILITY

Results behind the work

Professional outcomes connected to what matters: reducing manual effort, accelerating workflows and operating at scale.

daily NPSless time
nps_time
97%
~13 min → ~20 sec
Oracle validationless time
validation_time
81%
~3 min → ~34 sec
ZIP lookupseconds
zip_codes_cached
1M+
lookup: hours → seconds
processingmonthly scale
invoices_processed
5,000+
in a single month

Projects

All of them share the same core — processes, automation and data. Integrations, APIs, AI and software come in as tools to solve each specific problem.

01 / HUMAN-IN-THE-LOOP

Daily NPS

Focus query, history, comparison, PowerPoint generation and WhatsApp delivery in an assisted flow with closing and month-turn rules.

FOCUSHISTORYCOMPAREPPTWHATSAPPAPPROVAL
BEFORE~13 minAFTER~20 s≈ 97% less operational time
CASE FILE 02 / PROCESS VALIDATION SYSTEM
HUMAN DECISION / REQUIRED
CONTROLLED AUTOMATION · DOCUMENTS · ORACLE CRM

Oracle Validation

Human-in-the-loop application connecting Outlook and Oracle CRM On Demand, interpreting documents, applying blocking rules and verifying data persistence after write-back.

01ICMS / IPIclassification from document content
02IDENTITYCPF/CNPJ and date cross-checks
03RESILIENCEqueues and controlled failure recovery
04AUDITtraceability of human corrections
01OUTLOOK
02ORACLE
03PDFs
04RULES
05HUMAN
06VERIFY
AUTOMATION / CONTROLLEDFAILURE / RECOVERABLEWRITE / VERIFIED
03 / END-TO-END PIPELINE

Focus → Power BI

Python/Selenium pipeline automating extraction, SSO, publication of 4 datasets to SharePoint, and refresh of Dataflows and semantic models.

4AUTOMATED
DATASETS
FOCUSPYTHON / SELENIUMSSOSHAREPOINTDATAFLOWSPOWER BIPOWER AUTOMATE
// INDEPENDENT BUILDS

Systems built outside my day job

Complete products used to deepen backend, APIs, data, AI and software architecture skills.

More projects

All the code is open on GitHub
// 7 repositories · documented pipelines, backends, and models
view github ↗

Career Path

Started in operations and finance within the GM chain, then steered the route toward data and automation — where I work full-time today.

Operations
Data
Automation
Integrations
Systems
jun 2026 — present
Mid-Level Data Analyst
MSX InternationalMSX International·StellantisStellantis South America
  • Automated the Daily NPS workflow with human approval at the final step — Focus lookup, history, comparison, PowerPoint generation, monthly closing/rollover rules and WhatsApp delivery — reducing execution from ~13 minutes to ~20 seconds
  • Built an end-to-end Python/Selenium pipeline from Focus to Power BI Dataflow and semantic model refresh — extraction, SSO, publishing 4 datasets to SharePoint, and triggering via Power Automate
  • Rewrote DAX, Power Query and NLP across 4 corporate CX dashboards broken by a methodology change — dataflows and semantic model recovered without losing historical trend data
  • Proof of concept integrating Azure/Entra ID, MSAL and Microsoft Graph to update SharePoint files via an authenticated session
jun 2023 — feb 2026
Direct Sales Administrative Analyst
ConduentConduent·General MotorsGM Brasil
  • Built a human-in-the-loop Oracle CRM On Demand validation application integrating Outlook, process lookup, automated document analysis and final analyst approval — reducing execution from ~3 minutes to ~34 seconds
  • Implemented document intelligence and reliability controls for ICMS/IPI, CPF/CNPJ and dates, with automatic blocking rules, resilient queues, failure recovery, traceability and post-write validation in Oracle
  • Report consolidation via Python/VBA — 75% reduction in execution time
  • Management system with Outlook and SQLite cache of +1M ZIP codes77% less rework and lookups reduced from hours to seconds
jan 2023 — may 2023
Mid-Level Financial Analyst
ConduentConduent·General MotorsGM Brasil — Exportação & Logística
  • Built pivot tables and Excel formulas to speed up data analysis for the logistics team
  • Developed a VBA tool that queried the mainframe for real-time packaging status
jul 2022 — jan 2023
Junior Financial Analyst
ConduentConduent·General MotorsGM Brasil — Exportação & Logística
  • Helped implement a new triangulation process, speeding up operations and removing critical bottlenecks
  • Built a VBA macro that standardized logistics file formatting (PCK/LPN/ASN) before system submission, eliminating recurring character-count errors
apr 2021 — jul 2022
Administrative Assistant
ConduentConduent·General MotorsGM Brasil — Recebimento Fiscal
  • Processed +5,000 invoices in a single month, minimal error margin in SAP
  • Trained new hires on the fiscal receiving process
apr 2020 — apr 2021
Senior Process Operator
InfosysInfosys·General MotorsGM Brasil — Recebimento Fiscal
  • Document and invoice control; coordinated interface between departments
// STACK & CAPABILITIES

Tools are
means of execution.

I use the technology that solves the problem. Below is the ecosystem that appears most often across my projects and automations.

01AUTOMATIONprocesses & browser
Python
Python
Selenium
Selenium
VBA
VBA
Power Automate
Power Automate
02DATAprocessing & persistence
Pandas
Pandas
PostgreSQL
PostgreSQL
SQLite
SQLite
SQL
SQL
03BI & ANALYTICSmodels & decisions
Power BI
Power BI
Power Query
Power Query
DAX
DAX
ETLextract · transform · load
04SYSTEMSAPIs & applications
FastAPI
FastAPI
Docker
Docker
Git
Git
React
React
05INTELLIGENCEwhen the problem calls for it
Machine Learningmodels, classification and patterns
LLMsintegration and controlled generation
REST APIsintegration and orchestration

Behind the pipelines

Outside of code, I like stepping out of routine — travel and sports are part of that. It's in those moments away from the keyboard that most of my ideas actually show up.

SportsTravelFocusDiscipline

About

My work starts before the code: understanding the problem, researching the domain, mapping constraints and then designing a solution. That mindset connects my work in operations, reporting, internal automation and independent products.

Today I connect processes, automation, data, integrations and software: I turn manual routines into systems, validate assumptions with data and evolve projects as new needs appear. The focus is not only writing code, but building complete, useful and maintainable solutions.

Based inSanto André, SP — Brazil
Experience6+ years · GM / Stellantis chain
EducationData Science (in progress)
Independent projects7 independent builds
Current focusData & Automation
B.Sc. in Data Science
Cruzeiro do Sul (online) · São Paulo, Brazil
jun 2025 — jun 2027
Technical Diploma in Administration
Senac Santo André
jun 2018 — jun 2020