Business intelligence, commercial reporting, Power BI products and automation across international luxury and beauty operations.
Business intelligence · finance · AI
Business intelligence that survives contact with the business.
I turn messy commercial questions into data products people can trust: semantic models, Python pipelines, automation and AI-assisted workflows.
4+ yearsMexico · France · SpainBI / analytics / automation
4 → 20+automated reports
348stores connected
11automated QA checks
The thread between the work
I work close to the moment where information becomes action. Sometimes that means a Power BI model used every day by a commercial team. Sometimes it means documenting a room full of people building something new. In both cases, the job is the same: notice what matters, create structure and make it useful.
Explore the practice
Organisations
Where I have built experience.
The organisations are public. The operational details remain anonymised where the work is confidential.
Selected retail and commercial workflows within the wider luxury group context, described without exposing internal data.
Finance automation, ROI analysis and business analysis inside a large-scale retail environment.
CPFR, forecasting and supply-chain analytics across consumer-goods operations.
Choose your route
Two ways to work together.
A clearer entrance for people evaluating me for a role and people looking for a builder on a specific data or automation problem.
For hiring teamsLooking for a BI / AI builder?
See role fit, trajectory, organisations, leadership signals and the evidence behind the work.
Open recruiter profile ↗For project partnersHave a reporting or automation problem?See the systems I build: commercial BI, finance workflows, semantic models, QA and applied AI.
Explore selected projects ↗The story
I started by learning how to communicate. Then I learned how to make the system work.
The thread across my experience is not a specific tool. It is the habit of moving from a messy question to a clearer decision, and then building the structure that makes that decision repeatable.
Learn the room before modelling it.
Through communication, storytelling and photography, I learned to notice context: what people need to understand, what they are not saying yet and how a message changes when it has to move a group.
Follow the question to the process.
In finance, retail and supply chain, the work moved closer to the operational reality: forecasting, inventory, ROI, reporting and the definitions behind the KPI.
Turn repeated friction into a product.
Power BI models, Python pipelines, QA checks and documentation became the way I turn recurring manual work into something more reliable, reviewable and easier for a team to own.
Use new tools without losing accountability.
Today I explore Claude Code, MCP and LLM-assisted workflows with the same principle: the output should stay understandable, testable and connected to a real business decision.
Selected impact / by organisation
Different organisations. One operating pattern.
The work spans commercial BI, finance automation, supply-chain analytics and public developer tooling. The figures below stay separated by organisation and evidence level instead of being blended into one headline.
Luxury retail / Mexico + Iberia
From recurring reporting to reusable operating systems.
6documented commercial reporting pipelines across two brands, two countries and 5+ retailers
1,214–1,358 h/yrgross modeled capacity across four separately documented workflows; before overlap, review and adoption adjustments
4 → 20+reports scaled on a 25-table Power BI semantic model
114 / 235 / 25tables, measures and RLS roles documented in an enterprise model snapshot
177 / 2,181scoped EMEA contribution: documented measures and TMDL lines in a shared PBIP model
348 / 30 / 11stores, countries and automated QA checks in a separate worldwide sell-out workflow
The annual range is a modeled gross-capacity scenario from documented workflow baselines and cadences. It is not a realized consolidated saving: some workflows may share inputs, handoffs or analyst time.
Finance automation / large-scale retail
Turning finance questions into repeatable analysis.
ROIfinancial analysis and workflow automation inside a Digital CoE environment
Repeatablestructured inputs, reusable analysis and clearer handoffs for finance-related questions
Decision-readywork designed to connect financial interpretation with an operational next step
Public-facing framing stays qualitative until a reconciled before/after time baseline is available.
Supply chain / consumer goods
Making operational complexity measurable.
25k / 300+SKU-store combinations and points of sale analysed in supply-chain work
CPFRplanning, coordination and analytics foundations across commercial operations
Data → actionconnecting operational signals with planning and replenishment conversations
This is presented as supply-chain experience, separate from luxury-retail pipeline counts and time models.
Public repositories / technical evidence
Proof you can inspect, not just claims.
PBIP / TMDLpbip-doc documents Power BI projects with an AI-assisted workflow
Power BI visual engineeringtranspose-matrix-pro and FieldParameterSelector
SQL / data modellingsql-star-schema-northwind: star schema, 1,234 SQL lines and 12 quality checks
Applied toolingamazon-product-research-mcp: a TypeScript MCP server for structured product research
Only public repositories are linked. Client repositories, internal models and confidential source data remain anonymised or private.
Capability map
What I bring to the table.
A practical combination of business understanding, data craft and tools that make the result easier to trust and easier to run.
01 / Core
Business intelligence
Power BI · DAX · semantic models · KPI design · commercial reporting · forecasting
02 / Delivery
Data foundations
SQL · Python · ETL · data quality · automation · documentation · stakeholder adoption
03 / Applied
AI in the workflow
Claude Code · LLM APIs · MCP · prompt/context design · human review · reproducible tools
Selected work
What I have built.
Projects across luxury retail, consumer goods and large-scale retail, with a focus on business context rather than dashboards for their own sake.
From four reports to one operating model.
Scaled a Power BI product from 4 to 20+ automated reports on a 25-table semantic model. It became the single source of truth for Mexico revenue decisions.
Power BI · DAX · data modellingRead the case ↗Reporting that runs while the business moves.
Automated worldwide sell-out reporting across 348 stores and 30 countries with 11 data quality checks; the estimated manual baseline is ~30 minutes per run, while net monthly savings remain unmeasured.
Python · SQL · quality checksRead the case ↗A weekly commercial pack built around the decision.
Joined fragmented commercial inputs into a reviewable workflow for forecasting, replenishment and allocation; the documented comparison is ~4 hours of manual baseline versus ~2 minutes of run time plus ~10 minutes of review.
Python · forecasting · commercial BIRead the case ↗Stock reporting that makes exceptions visible.
Reconciled recurring stock inputs with reusable mappings and visible QA; the documented comparison is ~2–3 hours of manual baseline versus ~5 minutes, with review time still unverified.
Python · pandas · QARead the case ↗AI with a human still in the loop.
Built production pipelines and practical tools with Python, Claude Code and LLM APIs. The point is not to add AI everywhere. It is to make a real workflow clearer, faster and easier to own.
Python · Claude Code · LLM APIsRead the method ↗Some work is confidential. The evidence stays specific: scale, decisions, systems and the operational change behind them.
Good systems start with attention to the people using them.
How I work
Business first. Structure second. Tools where they help.
01Understand the decision
What needs to change, and who needs to trust the answer?
02Model the reality
One definition of each metric, with enough context to survive change.
03Automate the repeatable
Python, SQL, Power BI or AI workflows, chosen for the actual job.
04Make adoption possible
Validation, documentation and a product people can use every day.
In the room
The work is technical. The context is human.
From hackathons to commercial teams, the best systems start by paying attention to the people, language and constraints around the question.
A trajectory across places and disciplines
Mexico → France → Spain → Mexico
Business Intelligence at Tec de Monterrey, an MSc in International Management and Finance at Rennes School of Business, and more than four years working across international teams.
My current assignment with luxury retail / Iberia runs until December 2026. I am building toward the next chapter in Mexico City, around BI leadership, analytics engineering and practical AI.
Experience
The path behind the systems.
Different titles, one recurring thread: turning business complexity into something teams can use.
2026 → nowBI · luxury retail / Iberia
International commercial workflows, reporting automation and data quality. Building systems that remain useful across markets, teams and definitions.
2024–2026BI · luxury retail / Mexico
Power BI products, forecasting, sell-out reporting and Python automation for commercial decisions in a high-volume retail context.
2024Finance automation · large-scale retail
ROI analysis and workflow automation inside a Digital CoE environment, connecting financial questions with repeatable analysis.
EarlierCPFR · consumer goods
Planning, coordination and analytics foundations across commercial operations.
FoundationTec de Monterrey · Rennes SB
Business Intelligence, international management and finance — the academic base behind the technical and commercial perspective.
Public evidence
Repositories that make the craft inspectable.
Professional work stays anonymised. These public projects show the technical range behind the practice: semantic models, interfaces, SQL, MCP and finance tooling.
01 / PBIP documentationpbip-doc
AI-assisted documentation for Power BI projects, semantic models and TMDL/PBIR structure.
TypeScript · Power BI ↗02 / Visual engineeringtranspose-matrix-proA custom Power BI visual exploring how interaction and layout can make dense matrices easier to use.
TypeScript · Power BI ↗03 / Interaction designField Parameter SelectorA public visual experiment for selecting and ordering Power BI Field Parameters with persistent state.
TypeScript · UX ↗04 / Data foundationsSQL Star Schema NorthwindTwo fact-table grains, six business queries, 1,234 lines of SQL and 12 automated quality checks.
SQL · PostgreSQL · CI ↗05 / Applied AIAmazon Product Research MCPA TypeScript MCP server connecting structured product-research workflows with an agent interface.
TypeScript · MCP ↗06 / Finance systemsFinancial Planning PlatformA public-safe finance planning product exploring collaborative planning, scenarios and decision support.
Python · finance · product ↗What I write about
BI, AI-assisted workflows and the systems behind better business decisions.
If you work on data products, finance, automation or the transition from analysis to action, I would enjoy comparing notes.