Tech Lead Analytics Engineer

Transforming complex data into strategic insights through innovative engineering solutions

Years Experience
6 MBAs & Post-Grads Data Science, IT Project Management
80+ Technical Courses
3 Core Industries Finance, Logistics, Agribusiness
Scale
Startups to Enterprise Thrived in agile teams of 50 to global corps with 350k+ employees
Snowflake dbt Python R SQL Power BI Airflow BigQuery Git

About Me

I'm an Analytics Engineering Lead with over of experience transforming raw, complex data into strategic business intelligence. I don't just build pipelines; I build the systems of trust that let data speak for itself.

My career is built on designing robust data foundations, modernizing Kimball's dimensional modeling for the AI and data lake era using the Medallion architecture. I specialize in structuring enterprise data for classic BI, AI-consumption engines (like ThoughtSpot), and advanced Machine Learning features.

Domain Expertise:

  • Logistics & Supply Chain: Engineered data models to feed ML pipelines for predictive yard movements, driver allocation, revenue forecasting, and workforce optimization.
  • Finance & Risk: Built zero-to-one data marts for Collections and Fraud, enabling enterprise self-service in Tableau. Actively led the data discovery and modeling for ML models focused on fraudulent transaction detection and default prediction within aging buckets.
🏗️

Data Architecture

Designing enterprise-grade Snowflake, BigQuery, and Redshift data warehouses built to scale reliably

My technical focus spans the full analytics stack. I work fluently in Python, R, and SQL to tackle complex challenges — from credit card fraud detection with ensemble classifiers to financial risk analysis and operational intelligence. I believe precision is non-negotiable: the same obsessive attention to detail I bring to tuning a machine learning model is the discipline I apply when painting miniatures, mixing exact color recipes like Olive Green 545 with custom grey blends, because the final finish always matters.

🤖

Machine Learning

Building fraud detection and predictive models in Python and R, from feature engineering to production deployment

As a Tech Lead, I sit at the intersection of data engineering and business strategy. I bridge technical teams and executive stakeholders, translating data complexity into clear decisions and measurable outcomes. I've led data enablement programs across logistics, financial risk, and agtech — building cultures where data quality is owned, not assumed. My goal on every engagement is the same: leave the data landscape better, faster, and more trustworthy than I found it.

👥

Tech Leadership

Bridging engineering and business stakeholders to align data strategy with measurable organizational outcomes

I bridge the gap between business needs and technical execution across the E-commerce, Logistics, and Agribusiness sectors. I engage directly with B2C clients, B2B partners, and internal teams to gather complex requirements and feedback. My core strength lies in strategic triage: identifying whether a challenge requires a software feature, a data product, or a process improvement, and orchestrating the right teams to deliver high-impact solutions.

🤝

Stakeholder & Project Management

Triaging business challenges across sectors and orchestrating the right technical or process response to deliver measurable outcomes

Expertise

01

Data Architecture & Pipelines

Snowflake · dbt · Airflow · GitHub

→
02

Analytics & BI

Power BI · DAX · Domo · SQL

→
03

Leadership & Strategy

Tech Lead · COO · Agile · OKRs

→
04

Advanced Analytics & ML

Python · R · scikit-learn · XGBoost

→
05

Stakeholder & Project Management

B2C · B2B · Agile · Requirements

→
🏗️ Case Study — YMX Logistics

Data Architecture & Modern Pipelines

At YMX Logistics, I founded the data infrastructure from the ground up — starting from zero data maturity and building an enterprise-grade Snowflake environment with full governance, role-based access control, and automated quality monitoring. The entire stack was wired through dbt for transformation logic and GitHub for version-controlled, PR-reviewed data development.

Every pipeline was designed for idempotency and incremental loads. The architecture enabled reliable cross-organizational data sharing and self-service analytics for the first time in the company's history — eliminating the manual data pulls that had been standard before.

💡 Reduced data warehousing costs by 50% through query optimization and warehouse right-sizing, while simultaneously improving model coverage and pipeline reliability.
Tools Used
Snowflake dbt GitHub Actions SQL Airflow Python
📊 Expertise — BI Engineering

Analytics & Business Intelligence

Proven track record of building and scaling data ecosystems from scratch. At Smartbreeder, I founded the data department as the sole analyst, scaling it to a high-performing team of 12 by driving a company-wide shift in data culture and training. At CMA CGM, as the first dedicated analytics resource, I architected the end-to-end logistics data pipeline—from raw ingestion to Qlik Sense visualizations—enabling real-time executive decision-making. I specialize in transitioning organizations from siloed spreadsheets to governed, automated, and self-service environments.

💡 Engineered Power BI metadata dashboards at YMX to proactively detect data model integrity issues — turning reactive fire-fighting into systematic quality control.
Tools Used
Power BI DAX Domo SQL Snowflake BigQuery
👥 Case Study — Carbon (Agro-Industry Startup)

Leadership & Data Strategy

As COO at Carbon, a sugarcane agro-industry startup, I sat at the intersection of operations, technology, and executive decision-making. My role was to align the data and analytics roadmap with C-level business objectives — translating a fast-moving startup's operational reality into measurable KPIs and scalable data processes.

I led cross-functional teams where the data infrastructure had to keep pace with rapid product changes. This meant building governance frameworks that didn't slow down delivery — lightweight, iterative processes that gave stakeholders the visibility they needed without creating bureaucratic overhead that kills momentum in early-stage companies.

💡 Built the data culture from scratch in an environment with no prior analytics function — establishing the metrics framework that gave the leadership team reliable operational visibility for the first time.
Tools Used
OKR Frameworks Agile / Scrum dbt SQL Python BI Reporting
🤖 Project — Fraud Detection

Advanced Analytics & Machine Learning

One of my flagship ML projects involved building a credit card fraud detection system on a highly imbalanced real-world dataset (fraud rate < 0.2%). The challenge wasn't just model accuracy — it was precision/recall trade-off optimization where a false negative has direct financial cost and a false positive destroys customer trust.

I implemented an ensemble approach combining XGBoost, Random Forest, and Logistic Regression with SMOTE oversampling for class balancing. In R, I conducted the statistical validation and deep residual analysis. The final pipeline included automated model retraining triggers and a business-impact simulation layer that translated model performance into estimated financial risk reduction.

💡 The ensemble classifier achieved a 94%+ AUC-ROC and a precision-recall balance optimized for minimizing financial exposure — outperforming a baseline logistic model by 18 percentage points on recall.
Tools Used
Python R scikit-learn XGBoost SMOTE Pandas
🤝 Competency — Cross-Sector Client Engagement

Stakeholder & Project Management

Across E-commerce, Logistics, and Agribusiness engagements, I have consistently operated as the interface between business needs and technical execution. I engage directly with B2C clients, B2B partners, and internal teams to gather complex requirements, surface hidden assumptions, and translate them into actionable data or product specifications.

My core strength lies in strategic triage: identifying whether a challenge requires a software feature, a data product, or a process improvement, and then orchestrating the right technical and business teams to deliver. This role is part product manager, part solutions architect, and part account lead — a combination that accelerates delivery velocity across every engagement.

💡 At YMX Logistics, led cross-functional coordination across engineering, operations, and executive stakeholders — reducing miscommunication-driven rework and aligning sprint priorities to direct business value within the first quarter.
Tools & Methods
Agile / Scrum JIRA Confluence Requirements Gathering OKR Frameworks Stakeholder Mapping

Technical Skills

Platforms & Tools

Snowflake BigQuery Redshift Airflow dbt Fivetran

BI & Visualization

Power BI Domo Looker Tableau Data Studio

Languages & Databases

SQL Python R PostgreSQL MySQL

Cloud & DevOps

AWS GCP Azure Docker Git

Professional Experience

Data Analyst

YMX Logistics

Jul 2025 – Present

Logistics
Snowflake Power BI SQL dbt Python

Spearheading the data architecture and analytics foundation at YMX Logistics, owning end-to-end data infrastructure from Snowflake governance to Power BI operational monitoring. Serving as the central link between engineering, business stakeholders, and executive leadership to ensure data reliability and self-service adoption across the organization.

  • Founded the Snowflake data architecture from the ground up, implementing enterprise-grade governance, quality monitoring, and cross-organizational data sharing that enabled reliable data consumption at scale
  • Reduced data warehousing costs by 50% through SQL optimization and engineering best practices, resolving critical data problems that were directly impacting business operations
  • Engineered Power BI metadata dashboards to proactively investigate data model integrity issues, coordinating cross-functional reviews of missing data and reserve practices
  • Validated business metrics with stakeholders and drove self-service analytics adoption through structured user onboarding, documentation, and hands-on troubleshooting

Risk Data Analyst

WEX Inc

Nov 2023 – Jul 2025

Finance
Snowflake Tableau SQL Alation SCRUM

Delivered scalable Tableau analytics solutions and advanced SQL investigations within WEX's risk and financial data ecosystem, partnering with Data Engineering to build governed, reliable data marts. Served as a key bridge between business users and data infrastructure, championing governance and data quality across high-stakes financial domains.

  • Designed and maintained Tableau dashboards in close partnership with Data Engineering, building scalable data marts and documenting all solutions in Alation and Confluence for full governance traceability
  • Executed intermediate-to-advanced SQL analyses in Snowflake to investigate risk patterns, collections and receivables discrepancies, and fraud detection data problems
  • Facilitated data quality validation and Hypercare periods for new analytics initiatives, collaborating with business users to resolve data interpretation questions and ensure accuracy
  • Managed the analytics project roadmap through formal SCRUM ceremonies — Sprint Planning and Daily Stand-ups — coordinating priorities across cross-functional stakeholder groups

BPM Specialist

CMA CGM

Nov 2021 – Nov 2023

Shipping
QlikSense Lean Six Sigma KANBAN SQL

Led Digital Transformation and QlikSense analytics deployments for one of the world's largest container shipping companies, bridging data-driven insights with strategic business process improvement. Acted as the technical representative in executive-level meetings, translating complex operational data into actionable decisions.

  • Drove QlikSense analytics deployments as part of broader Digital Transformation programs, representing the technical team in strategic business meetings to align data capabilities with executive priorities
  • Developed and scaled best practices for analytics tool adoption and user communication, coordinating advanced training programs and cultivating an internal data community
  • Applied Lean Six Sigma methodologies to identify, quantify, and eliminate operational gaps and process inefficiencies, tracking and delivering improvements through a KANBAN framework

BI Analyst

SmartBreeder

Dec 2019 – Nov 2021

AgTech
SQL Server Power BI Python ETL

Owned the full BI stack for a fast-growing AgTech startup, designing SQL Server data warehouses and Power BI reporting layers from scratch while acting as the primary data enablement partner for both product and business teams.

  • Designed and maintained data warehouse structures in SQL Server, powering end-to-end Power BI reporting and proactively investigating data model gaps to ensure integrity across all business processes
  • Delivered intensive Power BI training programs to user groups across all skill levels — from business beginners to power users — providing ongoing support and troubleshooting to maximize adoption
  • Partnered with the product team to gather business requirements, investigate data-related questions, and facilitate User Acceptance Testing (UAT) for new analytics features

Projects Intern

Cargill

Jan 2019 – Dec 2019

Agribusiness
Power BI Power Apps VBA Excel

Contributed to digital product development at a global agribusiness leader, building internal tools to automate manual processes and improve reporting visibility across multiple departments.

  • Built custom internal tools using VBA (Excel), Power BI, and Power Apps, digitizing manual workflows for multiple internal departments and reducing operational overhead
  • Developed automated reporting solutions that improved data visibility and cut manual reporting effort for operations and logistics teams

Let's Work Together

I'm always interested in discussing data challenges, analytics strategy, and new opportunities. Feel free to reach out.

Email felipe.agro.terrazas@gmail.com
LinkedIn felipe-ramires-terrazas
Medium @felipe.ramires.terrazas
GitHub FelipeRamiresTerrazas

Send a Message

Resume / CV

Felipe Ramires Terrazas — Data Analyst

↓ Download CV

Portfolio Projects

End-to-end data science case studies spanning machine learning, time-series forecasting, and customer analytics.

Machine Learning

Credit Card Fraud Detection

Built a fraud detection pipeline on a heavily imbalanced real-world dataset (fraud rate < 0.2%). Benchmarked seven models and optimised the precision/recall trade-off to minimise missed fraud without unnecessary customer friction.

Python CatBoost Scikit-Learn SMOTE
Forecasting

Logistics Yard Moves Forecast

End-to-end forecasting pipeline for logistics yard operations. Pulls live data from Snowflake, trains a Prophet time-series model, and renders business-facing visual forecasts for operational planning.

Python Snowflake Prophet Matplotlib
Predictive Modeling

Customer Churn Prediction

Predictive churn model built in R using tidymodels with seven-dimensional hyperparameter tuning and Latin hypercube search. Surfaces the highest-risk segment — Month-to-Month + Fiber Optic — for proactive retention outreach.

R tidymodels Random Forest SQL
Regression

Food Delivery Time Prediction

Regression pipeline benchmarking seven models including a stacking ensemble. Converts raw GPS coordinates to haversine distance for geographic signal. SLA framing yields an actionable 93.5% promise-accuracy metric.

Python LightGBM Scikit-Learn Haversine
Customer Analytics

Customer Intelligence & Coupon Engine

Full customer intelligence pipeline: RFM segmentation, churn propensity scoring (ROC-AUC 0.992), and a prescriptive coupon engine that protects 66.9% of customers from unnecessary discount spend.

Python Pandas Random Forest RFM
Vendor Intelligence · Delivery Hero

Foodpanda Global Vendor Performance & Churn Risk

End-to-end pipeline across 11 Asian markets: consolidating 187K vendors and 3M reviews, cross-market EDA, and a Random Forest vendor churn-risk classifier (AUC 0.966) with a tiered intervention framework for ops teams.

Python scikit-learn Random Forest Multi-market