Senior Data Engineer/ Machine Learning/ Python Developer
Guillermo has over 4 years of experience as a Data Scientist and Software Engineer. During his career, he also served as the Chief Development Officer (CDO) and Co-Founder of a Video-Game Startup. His expertise is centred on Python, specifically Flask, Fast Api, Spark, PostgreSQL, and other SQL; he has also worked with TensorFlow, Keras, Scikit-Learn, PyTorch, PySpark, OpenCV, Django, Git, Docker, Kubernetes, MongoDB, AWS, and GCP. He recently led a Data team of five people, including himself. As a Co-Founder for his startup, he led and assisted his web development team of about ten developers, among other duties such as working as an unofficial product designer. He has worked with SCRUM and led projects using Agile methodology. Despite his primary expertise in data and backend development, he can also handle frontend development with Django and React.
Chief Data Officer | 7WISPS Web3 Studios
- Developed REST API using Python, FastAPI, SQLAlchemy ORM, Docker and PostgreSQL in AWS to serve 1M+ total requests.
- Architected SQL Operational DB from scratch using snowflake schema, serving 30k+ read and 2k+ write jobs daily.
- Integrated NoSQL stores from Web3 APIs (e.g. Moralis) using MongoDB and GraphQL to ingest 250k+ blockchain events.
- Deployed NLP content filtering for 300k+ community messages with RabbitMQ and GPT-3, eliminating 97%+ of toxicity.
- Configured OpenCV in Unity C# for object tracking, enabling ingestion of 12GB+ in-game data to OLAP Data Warehouse.
- Game Economy Design using Monte Carlo Simulations with Machinations API for collection of 5M+ match data results.
- Monitored all in-game economic data collection including 100k+ in-game item generations and inventory movements.
- Managed a 5 person Data Science & Econ department using Agile, with SCRUM methodology and Jira task management.
Data Scientist and Engineering Consultant
- Worked with top tier clients in Spain and UK to fine-tune and deploy OCR, CV and NLP-based document intelligence models.
- Developed image processing models for documents using Tesseract OCR and OpenCV to ingest 20+ GB of image data.
- Orchestrated the storage layer using Azure Blob Storage to handle the throughput of processed image data.
- Fine-tuned entity extraction for documents using NLTK and Python’s native RegEx to process the 20+ GB of image data.
- Containerized the model using Docker and orchestrated images with Kubernetes in CI/CD to reduce downtimes 60%.
- Advised clients on adoption of Big Data infrastructure, NoSQL vs SQL storage and identifying ML/AI use cases.
- Improved supervised learning model’s predictive power for Fraud Detection, raising ROC’s AUC score to 94%.
- Developed NNM recommendation engine for Spain’s #1 supermarket eStore’s Django backend, improving hit rate 28%.
Data Scientist/ Analyst
- Integrated unsupervised customer segmentation for user-personalized communications, increasing engagement 25%+.
- Analyzed promotional campaigns using A/B tests and automated KPI reporting e.g. customer retention, clickthrough, etc.
- Performed queries on SQL and Hadoop to source data for analyses and model of customer behavior with time series.
Data Scientist/ Analyst
- Built and owned optimization models for logistics workforce planning (resource efficiency), reducing waste by 18.8%.
- Implemented Holt-Winters time series model using numpy, pandas and statsmodels to forecast month sales at 2.2% MPE.
- Automated reporting of geolocation data mapping using matplotlib, enabling business strategy analyses for city growth.