Resume

Eleven years, most recently six and a half at AB InBev.

AB InBev - Data Analytics Lead, Logistics

Jun 2024 – Jun 2025

First holder of a new regional role bringing three countries' logistics data teams into one operation. Responsible for what got built, in what order, by whom, and whether the operation actually used it.

  1. I designed the operating model for a three-market Agile data team that had just been centralized and had no methodology yet: discovery, prioritization, the quarterly roadmap cycle, and the ceremonies themselves. I validated each piece with the three product owners.

    • Discovery ran per market because a process like warehouse picking carried different business rules in each one, so a combined session would have lost those differences
    • I extended weighted shortest job first scoring with multipliers for product type and market reach, which put the regional strategy directly into how initiatives were ranked
    • Result: each quarter the regional logistics director and the three market directors chose between two or three costed scenarios, each with a named resource plan by month. 91% of the committed roadmap delivered across the year

  2. I redesigned how dashboard usage was measured, interviewing sub-area managers to establish who should be using each report, at what frequency, and in which routine.

    • Every dashboard came out with an action attached: decommission, promote, fix, or leave. Low usage got investigated rather than assumed, and usually traced to something specific, like a source file owner who had stopped uploading
    • Result: over 85% usage in H2 2024. In 2025 I extended it across all three markets and built a portfolio rationalization method, presented to the regional logistics director, which identified the unused long tail for decommissioning

  3. I expanded the team's NPS survey from a small group of managers to the wider operation, building the form with conditional logic so questions adapted by market, sub-area, and which team the respondent worked with.

    • The wider base made the score meaningful, and the analysis could be read by market, by sub-area and by logistics pillar rather than only in aggregate
    • Result: 65.8 against a target of 60. For 2025 I unified the survey with the sister projects team and led the design for both areas

  4. I attended all three markets' daily stand-ups, which the role did not require, because most blockers raised there were organizational rather than technical: an unanswered question to a business partner, a missing permission, a file nobody had shared.

    • Sitting between the visualization analysts, the engineers and the business meant I could take those to the person who could actually clear them, instead of leaving them stuck inside one squad with no one who had the standing to move them
    • Result: the three squads started solving problems for each other. An engineer hitting something in one market got connected to whoever had already solved it in another, and I set up cross-market chat groups by role so that contact outlasted the stand-up

AB InBev - BEES Marketplace Finance and Strategy Lead

Jun 2022 – Jun 2024

Owned the finance, pricing and commercial reporting for a non-beer e-commerce business running inside the company's B2B ordering app, and built its financial and analytical infrastructure as the catalog grew from a handful of wine SKUs to over 500 products across six categories.

  1. I rebuilt the product master file as a proper data model, with separate dimension tables joined in Power Query and a calculation layer handling the unit conversions the catalog needed.

    • The output was a layer anyone on the team could use directly, with edit rights held on the source
    • Result: pricing, cost, logistics allocation and commercial tracking all fed from this one file

  2. I established how the ERP was actually calculating inventory cost, through diagnostic work with the procurement lead, then reproduced that logic in a model.

    • The system costed products through production recipes, which does not apply to goods bought and resold, and the logic behind its figures was not documented anywhere the team could reach
    • Result: a reliable cost per product, with procurement maintaining purchase data as the input layer. It became the cost basis for pricing decisions and monthly reporting

  3. I built the pricing methodology for the catalog, covering every product across every region and channel it sold in.

    • Many products had no market reference and no sales history, so the method had to produce a defensible price with nothing to compare against
    • Result: a tool that produced the full set of price scenarios per product. I taught Power Query and Power Pivot from scratch to my analyst, who moved up from an intern position into a finance and data analyst role and took pricing execution from there

  4. I built the commercial tracking model that became the team's daily reference, connecting sales exports, the cost model, the logistics allocation and the product master file through Power Query.

    • It tracked sales, discounting, revenue, order value and margin per product, cut by category, region, channel and partner
    • Result: used in every daily team meeting and the basis for monthly global reporting. A watchlist inside it flagged products that had slipped into negative margin, catching pricing errors in time to fix them live rather than at month end

AB InBev - BEES Rewards Lead

Jan – May 2022

Ran a country's B2B loyalty points program and rebuilt the targeting engine behind it, taking a manual pilot process into something that could run nationally.

  1. I rebuilt the targeting behind the program's challenge mechanic in Power Query, matching each customer to a cluster based on their own purchase behavior and setting a target scaled to it.

    • The inherited process was calculated by hand from flat Excel tables with manual lookups. Fine for a pilot in one channel, but there was no way to run a national base through it
    • Result: cluster granularity went from three to twenty, so the program got sharper as it scaled rather than blunter. Customers completing challenges grew from around 3,000 to around 9,000 across five months

  2. I designed four challenge objective types, each with its own audience, and matched assignment to customer profile.

    • Brand, sales and supply were all using the same mechanic for their own objectives, which risked a customer receiving overlapping challenges and accruing points twice on one purchase
    • Result: challenges were matched to each customer's profile, size and the brands they actually sold, and each function could use the mechanic without colliding with the others

  3. I extended the reporting that tracked whether challenges were actually being completed, joining purchase data from the ERP with challenge assignment data from the platform.

    • The two systems were not natively connected, and completion had to be measured per challenge, against only the products tied to that specific one
    • Result: visibility of the full funnel, from challenges sent through to completed, plus a watchlist of customers sitting on large point balances without redeeming

AB InBev - BEES Command Center Lead

Mar – Dec 2021

The analytics and BI resource for a country's commercial organization, and the only internal employee on a newly created centralized reporting function whose mandate was to own the company's sales information and become the single source of truth for the commercial teams.

  1. I built a self-service sales data product holding over four years of transaction-level history, at one row per client, SKU and day.

    • Historical sales data was being pulled in full from the source system each time anyone needed it, and that system capped history at three to five years
    • Result: history extended past the system's cap, published by distribution channel and distributed daily. It is still in use today, four years on

  2. I built a central product dimensions table covering brands, SKUs, sizes, segments and channels, and embedded it in that same model.

    • Product classification was held separately in each analyst's own file, so the same brand appeared under different labels and numbers from two teams did not reconcile
    • Result: correct classification applied automatically for every downstream user, rather than being maintained file by file

  3. I built the adoption reporting for the app's phased national rollout, tracking points of sale from eligible through registered, active, buying and fully digital, by region and by launch wave.

    • The app launched region by region across the year, so each wave had to be measured on its own before the next one went out
    • Result: I documented the full local event schema, mapping each in-app action to its event and properties, and benchmarked it against another market's implementation. I also maintained the cohorts that kept internal testing activity out of the adoption figures

  4. I built and maintained the Power BI reporting the commercial organization ran on: sales volume by region, channel, brand and SKU caliber, active buyer tracking with segmentation and period comparison, and the rewards dashboard for the loyalty program.

    • Each area had been producing its own reporting, so numbers arrived at meetings without agreeing with each other. Building it centrally meant one definition of a metric across every team that used it
    • Result: it became the source of truth for commercial performance across the organization. I later moved into the Rewards role and used the loyalty dashboards myself, from the other side

  5. I was the business-side interface for the contractor data engineering squad, working in an agile framework.

    • I translated commercial reporting requirements into specifications the engineers could act on, and explained the source system logic, ERP structures, business rules and channel definitions they had no context for
    • Result: I contributed to scoping which data was worth migrating to the internal datalake, based on what the business actually consumed, and wrote the onboarding material for new squad members

AB InBev - Sales Planning and Performance Manager

May 2019 – Mar 2021

The analytical and coordination support to a country Sales Director, embedded in the staff team as a one-person function, producing the numbers and the material that went into rooms up to regional president level.

  1. I took ownership of the monthly market research feed and turned it into processed analysis with the takeaways ready for each audience.

    • The data covered market share and distribution for both company and competitor products, but access to the raw files was restricted, so most of the commercial organization had no processed view of it. This sat outside my remit and I picked it up
    • I extended the same approach to two retail chain sell-out feeds that were reaching only the representatives covering those accounts, who were reading them for their own account rather than against the full history and the wider portfolio
    • Result: diagnostics segmented across five regions and four channels, with each regional sales manager receiving the cut for their own area rather than one general report. Processed market share visibility existed across the commercial organization for the first time, and I owned that workstream for the rest of the role

  2. I built a recurring system identifying points of sale performing worse than the rest of the base, scored on volume trend, purchase frequency and coverage.

    • Deciding which underperforming clients to prioritize had come down to intuition or one-off requests, with no way to tell whether the same ones kept slipping
    • Result: it refreshed on a cycle, so clients moved on and off the list as their performance changed rather than the list going stale. Regional sales managers used it to assign targeted follow-up, and it became part of the sales force's operating routine

  3. I built a daily volume tracking file consolidating multi-year sales history at one row per client, SKU and day.

    • History was re-downloaded in full from the source system each time anyone needed it, and that system capped what it held
    • Result: accumulated daily rather than re-pulled, so history built up past the cap. It was the direct predecessor of the self-service data product I built in the next role

  4. I supported a company-wide price structure reset, from pilot through national rollout.

    • The reset was commercially sensitive and rolled out in stages, so each stage needed evidence that it was working before the next one went ahead
    • Result: I analyzed volume and mix shifts between pilot and control points of sale by channel, and produced the presentations for the sales force and for leadership at each stage of the project

  5. I produced the sales content for every recurring executive forum the director took part in.

    • Where a forum needed more than one function's material, I consolidated inputs from Pricing, Route to Market and Trade Marketing into a single coherent presentation
    • Result: the analysis, the data and the narrative structure came from me and the director presented it, to audiences running from regional sales heads up to the regional Business Unit President

AB InBev - ZX Ventures Project Coordinator

Mar – Apr 2019

Two months of financial triage on two small consumer businesses inside the company's innovation arm, rebuilding the books from source records and documenting how the operation ran before handing it over.

  1. I rebuilt the P&L of a multi-location retail operation from source records.

    • The reported figures did not reconcile against invoices and inventory, so I worked back through invoices, purchase orders and contract documentation rather than from the existing reports
    • Result: a P&L that traced to source and a viability picture location by location, produced alongside the external accounting firm consolidating the statements

  2. I documented how both businesses ran and packaged it for the incoming team.

    • I reconstructed it from the available documentation and from a member of the previous team
    • Result: system tutorials for the order management ERP and the point-of-sale system, plus an organized archive of supplier accounts, credentials, contracts and open items

AB InBev - Pricing Coordinator

Dec 2018 – Feb 2019

Measured whether shops were selling beer at the prices the company recommended, and tracked whether distributors were meeting the sales targets their market plan payments depended on.

  1. I built the routine that measured price adherence from the monthly external survey files.

    • An external consultancy surveyed shelf prices across points of sale each month, but the surveyed prices and the suggested prices sat in separate places with nothing joining them, so there was no measure of whether the recommended structure was holding in the market
    • Result: a reusable template rather than one-off files, with a rolling archive extending back through inherited files, so adherence could be compared across periods rather than read one month at a time. Outputs went to the Pricing function and to commercial stakeholders by brand and region

  2. I tracked market plan compliance for the distributor and wholesale channels.

    • Payments under those agreements depended on hitting contracted volume and billing targets, which had to be calculated each period before payment could be released
    • Result: recurring monthly tracking of actual against contracted targets by SKU, with closing projections produced while the period was still open, so both sides could see where they stood in time to act on it

Euromonitor International - Market Research Analyst

Apr - Nov 2018

Country research for an international market research firm, working remotely and per project.

  1. I did the local fieldwork for a study on alcohol consumption in the country.

    • Reliable data on the market was hard to come by, so the picture had to be built from the ground up: storechecks recording what was actually on shelves, and interviews with the trade
    • Result: a well-rounded and impartial view of the market, built on observation rather than on what was already published

  2. I built and maintained the databases behind the study, then did the analysis and the reporting.

    • I ran more in-depth interviews with local actors than the study had reached in previous years
    • Result: a stronger evidence base under the firm's analysis and the findings delivered to the client

CADEP - Junior Researcher

Jul 2016 – Dec 2017

I coordinated fieldwork across two studies:

  1. I led a 5-person team administering the World Economic Forum's Executive Opinion Survey to more than 100 companies in Paraguay.

    • We got the highest response rate in three years
    • Result: the data was an input to the WEF Global Competitiveness Report, which helps investors understand the local landscape to drive better business decisions

  2. I oversaw the surveying of over 470 businesses as Paraguay's fieldwork for a four-country comparative research project on regional productive development and cooperation between firms (Chile, El Salvador, Paraguay and Uruguay), funded by Canada's International Development Research Centre and run across four national research institutions.

    • We surveyed more businesses than any other participating country
    • Result: the data was an input to an international policy seminar

  3. I taught myself STATA to automate the extraction of foreign trade statistics.

    • Result: the analysis could be re-run rather than rebuilt each time. I also conducted in-depth qualitative interviews with Paraguayan business owners on innovation in their sectors

FGV IBRE - Intern

May 2014 – Jan 2016

Economic research at one of Brazil's leading economic research institutions, on Brazilian foreign trade, Latin American economies, and the effect of China's rise on Brazilian trade.

  1. I built and maintained the databases behind the institute's recurring research, working from the aggregated databases of international organizations and Brazilian government entities.

    • Result: they were the evidence base the researcher I supported wrote her findings from, which went on to be published in the press and in academic papers

  2. I automated the generation of the statistics in Excel.

    • The figures were being compiled by hand each cycle, from sources that published on their own schedules
    • Result: the research could be produced on a cycle rather than rebuilt from scratch each time

Education

Bachelor's degree in Economics, UERJ, Rio de Janeiro.

Tools I work with