Skip to content
PT
Data Agents

Pedro Benther

I work in trade marketing. I code whatever slows me down.

I run the indirect pharma channel at a multinational and build what is missing to decide faster: automations, dashboards and AI agents.

contato@bnthr.org Résumé (PDF) LinkedIn GitHub

3 days → 5 min
Monthly incentive calculation, after automating it in Python.
~100 → 0
Governance findings per quarter. Zero for four quarters now.
R$ 30 to 50 M
In commercial incentives that I calculate and validate each year.
−95%
Less rework when closing rebates, thanks to automatic validations.

Experience

GSK

Three roles since June 2023, and what changed in each.

Since Feb 2025

Rio de Janeiro, hybrid

Indirect channel, national scope

Trade Marketing Analyst GSK Brazil, Vaccines and Medicines

I run trade for the indirect channel: 11 wholesalers, 7 focus brands and the calculation of incentives. The position was created because of my performance.

Calculation and governance

  • I automated the monthly incentive calculation in Python: from 3 days to 5 minutes.
  • I took governance findings from about 100 per quarter to zero, and kept zero for four quarters in a row.

Market intelligence

  • I built the weekly sell-in, sell-out and market share report from scratch. It is read from the board down to the product teams.
  • I merged the sell-out of 11 wholesalers into a single base, which gives an earlier read on demand before the official close.
  • 6 Power BI dashboards, used by 50 to 60 people across Trade, Marketing, Sales and Finance.
  • I calculate campaign ROI and prepare the numbers for the joint business plan and the forecast.

AI and field work

  • AI agents in production, using RAG over commercial data, for market analysis and documentation.
  • A price-capture app for the point of sale: 18 field users, about 1,000 records a month.

If I did it twice by hand, the third time it is a script.

Feb 2024 to Feb 2025

Rio de Janeiro, hybrid

Direct channel, about 100 clinics

Trade Marketing Intern GSK Brazil, Vaccines

I fixed rebate management and produced the market analyses the board used to decide on discounts.

  • I restructured rebate management with automatic validations: 95% less rework at close.
  • My market analyses supported a board decision with an estimated saving of £5 million.
  • The automations and documentation rules I left behind were adopted by the area's governance.

Jun 2023 to Feb 2024

Rio de Janeiro, on site

Marketing in a regulated environment

Marketing Intern GSK Brazil, Antibiotics and CNS

Where I learned the process before the data: promotional material, compliance and my first analyses.

  • I coordinated approval of promotional materials with the Medical and Compliance teams.
  • I helped design the omnichannel strategy and the portfolio's communication cadence.
  • I ran performance and market analyses for the Marketing manager.

Education: Marketing, Universidade Veiga de Almeida (2025). Languages: Portuguese (native), English (advanced, C1).

Projects

The ones with a demo you can try right here, with made-up data.

Data

Agents

Open source

Tools

What I use day to day.

Reading the market
Sell-in and sell-out, market share, campaign ROI, incentive calculation, joint business plan, forecast, IQVIA
Handling data
Python, SQL, Power Query, advanced Excel, VBA
Showing it
Power BI, DAX, Microsoft Fabric, React, ECharts
Automating
AI agents, RAG, MCP, Claude Code, Power Apps, Power Automate