Transforming data into decisions, building intelligent systems that scale
I lead data-driven innovation at Mercado Libre (Latin America's largest tech unicorn), where I build intelligent systems that democratize commerce and financial services for millions of users.
class MarianoGobeaAlcoba:
def __init__(self):
self.role = "Data & Analytics Tech Leader @ Mercado Libre"
self.location = "Buenos Aires, Argentina 🇦🇷"
self.team_size = "+6 data analysts & scientists"
self.focus = ["Security BI", "AI/ML Engineering", "Backend Development"]
def current_impact(self):
return {
"ai_adoption": "+30% across key areas",
"operational_efficiency": "+20% improvement",
"analysis_time_reduction": "40% faster with predictive models",
"data_errors_reduction": "70% fewer critical errors",
"annual_business_impact": "+USD 500K estimated",
"team_engagement": "+25 points eNPS increase"
}
def recent_achievements(self):
return [
"🔬 Built unified Forensic API (Python + Flask) for Security products",
"🤖 Designed MCP ecosystem for LLM-powered forensic analysis",
"📊 Architected RAG systems (+35% contextual accuracy)",
"🚀 Deployed fine-tuned BERT & GPT models (+20% automation)",
"⚡ Accelerated time-to-market by 50% with MLOps practices",
"🔧 Automated workflows with Airflow, dbt, n8n (+60% efficiency)"
]Tech Leader • Sr Data & Analytics Engineer • Product Owner
- 👥 Leading team of +6 data professionals (analysts & scientists)
- 📈 Increased AI adoption by 30% across critical business areas
- 🚀 Improved operational efficiency by 20% through strategic alignment
- 💪 Elevated team engagement: +25 eNPS points via mentoring & agile practices
Security & Forensics:
- 🔐 Designed & deployed Forensic API (Python + Flask) unifying 8 security products
- 🤖 Built MCP (Model Context Protocol) ecosystem for LLM-powered investigations
- ⚡ Reduced forensic analysis time by 70-80% with AI-assisted tools
- 🔍 Integrated with Cursor IDE, Meli-GPT, and Verdi Flows for automated forensics
AI/ML Engineering:
- 🧠 Architected RAG systems improving LLM contextual accuracy by 35%
- 🎯 Trained & deployed fine-tuned BERT & GPT models (+20% automation)
- 📊 Built ML-ready pipelines with scikit-learn & custom transformers
- 🚀 Accelerated time-to-market by 50% with MLOps adoption
Data Engineering:
- 🏗️ Designed scalable architectures handling massive volumes (45% better ETL performance)
- 📦 Migrated authentication model from AWS Athena → Google BigQuery
- 🔄 Automated workflows with Airflow, dbt, n8n, Zapier (60% efficiency gain)
- 📊 Built data models for Login, Reauth, Factors (MP, ML, ME platforms)
Data Analytics:
- 📊 Developed strategic dashboards (conversion, auth factors, account recovery)
- 🎯 Generated actionable insights reducing analysis time by 40%
- 💡 Established robust data quality processes (70% reduction in critical errors)
- 📈 Created executive-level reporting trusted by leadership
UADE (Universidad Argentina de la Empresa)
- 📚 Adjunct Professor - Programming Workshop I
- 🎯 Teaching first-year students foundational programming concepts
- 💡 Designing modern curricula with industry best practices
Soy Henry (Leading Latam Coding Bootcamp)
- 🚀 Instructor - Python, SQL, Data Analytics
- 👥 Mentoring future tech professionals
- 🌟 Contributing to tech talent democratization in Latin America
Problem: Manual forensic analysis was slow, error-prone, and didn't scale
Solution: Built unified API + MCP ecosystem + AI-powered tools
Impact:
- ⚡ 70-80% reduction in forensic investigation time
- 🤖 AI-assisted analysis via Cursor, Meli-GPT, Verdi Flows
- 🔗 Multi-product correlation (Login, Reauth, Factors, ITO, Recovery, etc.)
- 📊 Real-time dashboards for security teams
Tech Stack: Python, Flask, BigQuery, MCP, LLMs, n8n, Slack API
Problem: No unified view of authentication metrics across products
Solution: End-to-end data platform with models, dashboards, and automation
Impact:
- 📊 Centralized analytics for 8+ security products
- 📈 Strategic dashboards used by executives and product teams
- ⚡ Automated daily updates reducing manual work by 90%
- 🎯 Actionable insights for UX optimization and fraud prevention
Tech Stack: BigQuery, Looker Studio, Python, SQL, Dataflow
Problem: DataMesh alerts were noisy, duplicated, hard to interpret
Solution: Intelligent consolidation agent using GenAI
Impact:
- 🔔 Reduced alert noise via smart consolidation
- 🤖 AI-generated summaries for faster incident response
- ⚡ Automated routing to Slack with context
Tech Stack: n8n, Verdi (GenAI), Slack API, Python
technical_depth:
- "Full-stack data professional: engineering + analytics + ML"
- "Polyglot: Python, R, Go, Java, Kotlin, SQL"
- "Cloud-native: GCP (BigQuery, Dataflow, Cloud Functions)"
- "AI/ML: From sklearn to LLMs, from training to production"
strategic_thinking:
- "Background in sociology: understanding human behavior & systems"
- "Product management experience: business acumen + technical execution"
- "Leadership: building high-performing teams with +25 eNPS"
impact_orientation:
- "Everything measured: ROI, metrics, business outcomes"
- "USD 500K+ annual business impact from data initiatives"
- "Focus on automation: 60-90% time savings in manual processes"
continuous_learning:
- "Currently: Postgraduate in Software Engineering (UAI)"
- "Teaching: Giving back by educating next-gen talent"
- "Always exploring: MLOps, LLMs, vector DBs, RAG architectures"|
|
|
|
|
|
|
|
| Metric | Impact | Context |
|---|---|---|
| 🤖 AI Adoption | +30% | Across key business areas |
| ⚡ Operational Efficiency | +20% | Strategic alignment with corp objectives |
| 📊 Analysis Speed | -40% reduction | Via predictive models & automation |
| 🎯 Data Quality | -70% errors | Robust data quality processes |
| 🚀 Time-to-Market | -50% faster | Emerging tech adoption (MLOps, LLMs) |
| 💰 Business Value | +USD 500K | Annual estimated impact from initiatives |
| 👥 Team Engagement | +25 eNPS | High-performance culture via mentoring |
| 🔍 Forensic Analysis | -70-80% time | AI-powered forensic suite |
| 🔧 Workflow Automation | -60% manual work | Airflow, dbt, n8n, Zapier |
| 🧠 RAG Precision | +35% | Improved LLM contextual responses |
| 🏗️ ETL Performance | +45% | Scalable architecture for big data |
The Challenge:
Security teams at Mercado Libre needed to investigate authentication incidents across 8+ products (Login, Reauth, Factors, ITO, Recovery, etc.). Manual forensics took hours and lacked cross-product correlation.
The Solution:
- 🏗️ Built unified Forensic API (Python + Flask) aggregating 8 security products
- 🤖 Designed MCP (Model Context Protocol) server for LLM-powered investigations
- 🔗 Integrated with Cursor IDE, Meli-GPT, and Verdi Flows (n8n)
- ⚡ Deployed serverless architecture (low latency, minimal downtime)
The Impact:
- ⚡ 70-80% reduction in investigation time
- 🔍 Cross-product correlation in seconds (previously: manual, hours)
- 🤖 AI-assisted forensics via natural language queries
- 📧 Automated forensic reports via email/Slack
Tech: Python, Flask, BigQuery, FastMCP, LangChain, n8n, Slack API
The Challenge:
No unified analytics for authentication flows across Mercado Libre & Mercado Pago ecosystems.
The Solution:
- 🏗️ Designed & built end-to-end data models for Login, Reauth, Factors
- 📊 Created strategic dashboards for conversion, auth factors, recovery
- ⚡ Automated daily updates via efficient ETL pipelines
- 🎯 Generated actionable insights for product teams
The Impact:
- 📈 Centralized analytics used by executives & product managers
- ⚡ 90% reduction in manual reporting work
- 🎯 Data-driven decisions for UX optimization
- 💰 Cost optimization in data operations
Tech: BigQuery, Looker Studio, Dataflow, Python, SQL
The Challenge:
DataMesh alerts were noisy, duplicated, and hard to interpret—leading to alert fatigue.
The Solution:
- 🤖 Built AI-powered consolidation agent using GenAI
- 📝 Generated smart summaries from multiple alerts
- 🔔 Automated routing to Slack with context
The Impact:
- 🔕 Reduced alert noise via intelligent deduplication
- 🤖 AI summaries for faster incident interpretation
- ⏱️ Faster response time from engineering teams
Tech: n8n, Verdi (GenAI), Slack API, Python, BigQuery
The Challenge:
Regional logistics operations lacked standardized KPIs and monitoring systems.
The Solution:
- 📊 Standardized regional KPIs across countries
- 🎯 Implemented operational monitoring for critical indicators
- 🚀 Led continuous improvement initiatives in cross-docking
The Impact:
- ⏱️ Reduced processing times in MLA logistics
- 💰 Cost optimization in cross-docking operations
- 📈 Improved customer experience at critical journey points
Tech: Excel, Power BI, Python, SQL Server
🎓 Postgraduate in Software Engineering
Universidad Abierta Interamericana (UAI) | 2025-2026
🎓 Bachelor's Degree in IT Management
UADE | 2020-2021
🎓 Teaching Degree in Sociology
Universidad de Buenos Aires (UBA) | 2014-2015
🎓 Bachelor's Degree in Sociology
Universidad de Buenos Aires (UBA) | 2007-2014
📜 Python Backend Engineer Certificate
Platzi | 2022-2023
🏆 Multiple Security & Development Certifications
✓ Secure Backend Development (Python)
✓ Product Owner Fundamentals
✓ Emotional Intelligence
| Data Engineering | AI/ML | Leadership | Backend Dev |
|---|---|---|---|
| ETL/ELT Pipelines | RAG Architectures | Team Leadership (+6) | Python/Flask APIs |
| Data Modeling | Fine-tuned LLMs | Agile Methodologies | REST/GraphQL |
| BigQuery Expert | scikit-learn | Strategic Planning | Serverless (GCP) |
| Airflow/dbt | PyTorch/BERT | Mentoring & Upskilling | Microservices |
| Data Quality | Vector DBs | Stakeholder Mgmt | CI/CD Pipelines |
Check out my GitHub Profile for:
- 📦 100+ repositories (data analytics, backend, ML projects)
- 🔀 Active contributions to open-source
- 📚 Educational repos for teaching Python, SQL, and data analytics
- 🚀 Production-grade code from Mercado Libre projects
- 🔬 Building MLOps pipelines for authentication security
- 🤖 Exploring agentic AI workflows with LangChain & CrewAI
- 📚 Teaching Python & Data Analytics at UADE & Soy Henry
- 🎓 Completing Postgraduate in Software Engineering at UAI
- 🚀 Leading AI adoption initiatives in Security BI team
I'm always open to:
- 🤝 Collaborating on data engineering or AI/ML projects
- 💡 Discussing authentication security, fraud prevention, or LLM applications
- 🎓 Speaking about data analytics, MLOps, or tech leadership
- 👥 Mentoring aspiring data professionals and backend developers
💡 Open to opportunities in Data Engineering, AI/ML Engineering, or Tech Leadership roles
"Democratizing commerce and financial services through data, one insight at a time"
Working at Mercado Libre • Teaching at UADE • Building the future of AI-powered analytics