Andrés Fernández

aka Jorge, Jorge Andrés, Fernández Jiménez, or just fernandrez

Machine Learning Engineer & Data Scientist

A passionate technologist with over 13 years of experience transforming complex data into actionable intelligence. I bridge the gap between raw information and strategic insights, helping organizations harness the power of their data to drive innovation and sustainable growth.

With expertise spanning machine learning, natural language processing, computer vision, and big data architecture, I thrive in environments that challenge conventional thinking. My career has been built on turning abstract possibilities into tangible solutions that create measurable impact.

Currently focused on large language models, multimodal systems, and knowledge graph applications that blend cutting-edge AI with practical business value. I believe that at the intersection of technology and human creativity lies the potential to solve our most pressing challenges—both for business and society.

Technical Skills

AI & Machine Learning

Transformer Models Large Language Models Vision Transformers Multimodal AI Document AI Neural Networks Reinforcement Learning Knowledge Graphs PEFT/LoRA

NLP & Computer Vision

SparkNLP Hugging Face NLTK spaCy OpenAI Vision Qwen-VL OCR Document Processing Semantic Search

Languages & Frameworks

Python Scala C# JavaScript PyTorch Scikit-learn PySpark

Data & Cloud

Apache Spark SQL/NoSQL Hadoop AWS GCP Azure Docker Kubernetes MLOps

Latest Focus Areas

Currently exploring and implementing solutions using:

Document Vision Transformers (DoVT) Multimodal Fine-tuning Retrieval Augmented Generation (RAG) LLM Agents Knowledge Graph Integration with LLMs Enterprise AI Architectures

Initiatives

iSeeCI

Corporate Intelligence Strategies

Data-driven corporate intelligence solutions providing strategic insights for businesses. Leveraging AI to enhance competitive analysis, market positioning, and decision-making processes.

Tailoredia

Customized Intelligence Solutions

Bespoke artificial intelligence solutions designed for specific business needs, bringing data to life. Specialized in enterprise-grade NLP and knowledge graph applications for complex document processing.

Projects

DeepScope

Manifold Visualization

Low scale app to learn, process and visualize manifolds generated by applying vectorization / embeddings techniques to datasets. Finds outliers, targets them and take action on them to improve all your datasets.

Python Embeddings UMAP t-SNE Outlier Detection

AgenTAI

Semantics Aware Agent

Ontology driven Chatbot supporting Sockets and Telegram. Provides you with the necessary project structure and barebones functionality to develop your semantic aware agent using FastAPI, Sockets and Pydantic.

Python FastAPI Ontologies Semantic Web Chatbots

Tailored Media

Comprehensive Image Processing & Analysis Pipeline

Advanced image processing application that performs multiple AI-powered analyses including metadata extraction, object detection, classification, segmentation, and scene description. The system presents all results in an interactive map view with sophisticated filtering capabilities for metadata and detected objects/scenes.

Computer Vision Object Detection Image Segmentation Scene Understanding Metadata Analysis Geospatial Visualization

Experience

John Snow Labs

Machine Learning Engineer

Jun 2019 - Present · 5 yrs 10 mos

Specialize in Natural Language Processing (NLP) to process documents and extract valuable information in the form of relation graphs for different document ontologies. Implement SOTA NLP models in Scala / Spark / Python. Tune, ensemble and customize NLP models for client engagements and lead the Data Science Forum for team knowledge sharing.

Mondra

Data Scientist

Oct 2024 - Feb 2025 · 5 mos

London, UK

Optimized time and performance in a machine learning model for prediction of carbon emissions based on food recipes. Reduced inference time by 40% while maintaining prediction accuracy through algorithm refinement and feature engineering.

Combine AI

ML Engineer / Software Architect

Jan 2023 - Sep 2024 · 1 yr 9 mos

London, UK

Developed a full end-to-end RAG system for personalized meeting summarization for the Eric.ai project. Implemented using open source LLMs (Mistral and Llama) with async/await pattern and Celery task framework on Redis. Managed the complete software lifecycle including repository management and Kubernetes deployments.

EY

Assistant Director in Client Technology Platform

Aug 2018 - Nov 2019 · 1 yr 4 mos

Greater Buenos Aires, Argentina

Led a team of two to design, implement and support cloud platform architecture monitoring on Azure. Designed and implemented time series analytics for adaptive thresholds in alert rules and ML pipeline to process terabytes of event logs using Databricks.

AlixPartners

Senior Vicepresident

Jun 2016 - Jul 2018 · 2 yrs 2 mos

Greater Buenos Aires, Argentina

Led a team standardizing data processes for Digital Marketing using Python, SQL Server, .NET, NodeJS. Designed data integration pipelines and ETLs, created algorithms for customer normalization, and built ML pipelines for NLP in procurement.

CRISIL / Wells Fargo

Data Science Lead for Wells Fargo CMoR

Dec 2015 - Jul 2016 · 7 mos

Designed and implemented Big Data Analytics and Visualization Pipeline using Scala, SparkML, Hadoop, NodeJS, and D3.js. Built big data web services with Scala and AKKA, and visualization services with NodeJS.

iSee Corporate Intelligence

Director of Finance

Sep 2014 - May 2016 · 1 yr 9 mos

Data Driven Decision Making (DDDM) using Time Series models and Machine Learning. Tracked and controlled business budget execution and financial operations.

Mozido

Data Engineer for Customers JCCUL and Dairy Queen

Sep 2013 - Dec 2015 · 2 yrs 3 mos

Conceptualized, Designed, Built and Implemented the Payment Reporting Architecture using XML, PostgreSQL, Java and Pentaho. Designed and developed ETLs for creating a Reporting Datamart of Customer Payments grouped by different branches based on Government Requirements. Created Data Access and Business Logic Layers in an XML – Java – PostgreSQL Architecture for Report and Analytics distribution over the Project Stakeholders, using the previous Datamart.

Hipercept Solutions Colombia

Data Engineer for Customers Jamestown and Brookfield

Jan 2013 - Jun 2013 · 6 mos

Medellín, Colombia

Modeled, Simulated and Analyzed Real Estate Assets and Funds operations using specialized software to optimize the involved processes. Developed a Stochastic Model to represent the "random" outcomes of contract renewals based on Tenant Risk metrics (PHP-MySQL). Created RESTful Web Service Architecture to manage and analyze Lease contracts on registered properties. Built Commercial Web App using the previous middleware to present information and charts on Leased Properties.

Mapir Research Group

Social Robotics Researcher

Jan 2012 - Sep 2012 · 9 mos

Málaga, Spain

Conducted research activities focused on development of Social Enabled robots, specifically in audio interaction (ultimately conversation). Developed various algorithms: one to filter audio streams and direct servos to sound sources, another to ponder sound source angles based on intensity, and a third to direct mechanisms with two rotational degrees of freedom using "Respect" and "Speed" parameters. Created an algorithm to estimate these parameters based on a personality array containing values for Psychosis, Extroversion and Neurosis (PEN model by Eysenck). Implemented all projects as modules for Open Source Robotics Libraries MRPT and OpenMora in C++.

Suramericana Insurance (SURA)

Actuarial Analyst

Dec 2008 - Sep 2011 · 2 yrs 10 mos

Medellín, Colombia

Performed Data Mining and Intelligent Systems Modeling for Insurance Business including Reserve Models, Pricing Models, and Loss Models. Developed an Adaptive Inference System for casualty insurance pricing, based on the underlying stochastic models for each type of insurance (Automobile, Fire, Life and Health) using ASP.NET and SQL Server. Created a RESTful Web Service Architecture to manage Automobile Insurance Pricing through the parametrization of a Generalized Linear Model, involving numerous meetings with business executives and detailed requirements gathering to successfully parametrize the GLM.

Education

Universidad de Málaga

MSc. Software Engineering and Artificial Intelligence

Oct 2011 - Oct 2012

Research: Detection and orientation of a mobile robot towards a sound source using Neural Networks.

Universidad Nacional de Colombia

Actuarial Science for Insurance Companies

Jan 2009 - Mar 2010

Project: Loss model for hedge funds, using diverse random variable generation algorithms.

Escuela de Ingeniería de Antioquia

Mechatronic Engineering

Jul 2004 - Dec 2009

Research: Actuarial Management Platform using Supervised and Unsupervised Neural Networks.

Contact

Personal Email

iSeeCI Email

Tailoredia Email