I build AI automations, data pipelines, and BI dashboards for international trade and business operations, turning sourcing, customs, and logistics challenges into measurable business results.

🧭 Quick navigation | Click a section
About | Achievements | Tech Stack | Projects | Activity | Contact

I am an Argentine 🤖 Data Analytics & AI Automation Engineer and 🎓 B.A. in International Trade with Polish citizenship. Working in international trade since 2018, I combine hands-on business experience with AI, data engineering, and analytics to automate operations and build commercial products.

• Advanced English: Extensive experience leading international negotiations and collaborating with multicultural teams (USA, China, India, Malaysia, Brazil).
• Native Spanish: Born and raised in Argentina.
• Basic Chinese: Began learning in 2018; applied to technical documentation and commercial relationships.
• EU Citizen: Open to Remote/Relocation opportunities.

_{Logos are trademarks of their respective owners, shown only to identify employers, clients and education. No endorsement is implied.}

• Multi-Agent AI Sourcing SaaS: Built an internal sourcing system and commercialized it as a SaaS offering operating 24/7. Its 10+ AI agents find and negotiate with Chinese suppliers, support product classification through RAG, and generate interactive landed-cost quotations, while a web app and dashboard track key metrics, including costs. Since June 2026 it has contacted about 570 suppliers and secured about 220 quotations a month, at a median price 11% below the supplier's own listed price. AI writes 99.9% of the negotiation messages, 71% of them outside Argentine business hours while China is at work, with AI costs under USD 0.10 per quotation: the work of more than two people for under 2% of the cost of one entry-level employee in Argentina (estimate).
• Mercado Libre Customer Q&A Bot: Answers buyer questions 24/7, checking every five minutes, with 97% answered without human intervention and a median response time of 10 minutes. 51% of questions arrive at night or on weekends and still get answered. It costs 95.8% less than the previous tool, under USD 0.50 a month in AI, and a dashboard tracks its key metrics, including costs.
• AI Cost Optimization: Cut the monthly AI spend of the international negotiation agents by 96% by optimizing model usage, splitting agents into modules and adding validation controls, while keeping 95%+ hand-reviewed accuracy and fixing errors the previous setup made in production. A simple, customizable presentation builder in the web app also replaced AI-generated presentations, cutting that cost to almost zero.
• AI-Powered Macroeconomic Reporting: Built a fully automated, cloud-hosted pipeline tracking FX rates, country risk, interest rates, inflation, monetary aggregates, public and private debt, and Bitcoin. Every business day it emails seven charts with AI conclusions and a nine-slide executive deck, with no human intervention and zero running cost: Python computes the indicators; Gemini writes the conclusions. A Cloudflare Worker triggers each run on GitHub Actions, both on free plans, while automated validation, duplicate-run protection, failure alerts, and 260+ offline tests keep delivery reliable.
• Self-Service Analytics for Microsoft: Reduced operational workload for Microsoft's global sales team by deploying a self-service Power BI solution, eliminating manual email-based queries (100% English-led role).

Category	Tools
Ingestion & ETL
BI & Viz
Storage & Data Ops
Automation
AI APIs

Selected projects showing my analytical approach, business impact, and data visualization.

From market data to executive reporting, without manual compilation. Tracks Argentine FX rates, country risk, BCRA and Federal Reserve interest rates, inflation, monetary aggregates, public and private debt, and Bitcoin. Every business day at 16:00 Argentina time, the pipeline validates and stores the data in Supabase and emails seven charts with AI conclusions plus a nine-slide PowerPoint deck, with no human intervention and zero running cost. Python computes the figures; Gemini writes the conclusions. The latest charts and deck are published automatically to the repository.

• Reliability: Concurrent ingestion, schema validation, duplicate-run protection, AI model failover, failure alerts, and a daily completeness check. 260+ offline tests run in CI.
• Stack: Python, Pandas, Playwright, httpx, Pydantic, Supabase, Gemini API, Jinja2, Matplotlib, Seaborn, python-pptx, Cloudflare Workers, GitHub Actions, pytest.
• Open source (Apache 2.0): The full source code is public, so anyone can replicate the pipeline with their own keys (see .env.example) on free plans.

Daily executive deck, built and published automatically by the pipeline.
▶️ View today's deck online

Market share analysis and KPI visualization for potential importers across Latin America. Designed to identify strategic partners and competitive landscape shifts.

• Stack: Power BI, DAX, Power Query, Data Modeling.

• 🌐 Website: www.globalaize.com
• 💼 LinkedIn: /in/mauroemartinez
• 📧 Email: mauro@globalaize.com
• 📱 WhatsApp: +54 9 11 5990-9647
• 🌍 Availability: Open to 100% Remote (Worldwide) or Relocation.
• ⚖ Legal: Argentine & EU Passport Ready (No sponsorship required for EU-based roles).
• 📍 Current Location: Buenos Aires, Argentina.

"AI & data to solve global trade problems."