Full Stack Developer

• Oct 2023 - Present

Data Scientist

• Mar 2023 - Sep 2023

Design and development of an automatic text processing application NLP NER by Deep Learning methods: . Automate the labeling of a dataset via GPT generative models to build a training dataset. . Use an OCR model (textract/tesseract) to transform the document into raw text. . Train HuggingFace/Spacy models to perform Named Entity Recognition (NER) tasks. . Manage the life cycle of the created models and optimize them via MLflow/AutoKeras. . Deploy the solution in a web application… Show more Design and development of an automatic text processing application NLP NER by Deep Learning methods: . Automate the labeling of a dataset via GPT generative models to build a training dataset. . Use an OCR model (textract/tesseract) to transform the document into raw text. . Train HuggingFace/Spacy models to perform Named Entity Recognition (NER) tasks. . Manage the life cycle of the created models and optimize them via MLflow/AutoKeras. . Deploy the solution in a web application using Streamlit/FastAPI. Show less Design and development of an automatic text processing application NLP NER by Deep Learning methods: . Automate the labeling of a dataset via GPT generative models to build a training dataset. . Use an OCR model (textract/tesseract) to transform the document into raw text. . Train HuggingFace/Spacy models to perform Named Entity Recognition (NER) tasks. . Manage the life cycle of the created models and optimize them via MLflow/AutoKeras. . Deploy the solution in a web application… Show more Design and development of an automatic text processing application NLP NER by Deep Learning methods: . Automate the labeling of a dataset via GPT generative models to build a training dataset. . Use an OCR model (textract/tesseract) to transform the document into raw text. . Train HuggingFace/Spacy models to perform Named Entity Recognition (NER) tasks. . Manage the life cycle of the created models and optimize them via MLflow/AutoKeras. . Deploy the solution in a web application using Streamlit/FastAPI. Show less

AI Engineer

• Oct 2022 - Dec 2022

Processing leads signal in order to automate ECG Classifying : . Exploratory Data Analysis: extract the metadata from ECG images using EasyOCR and Keras, gather patients informations and have an idea about the ECG disease distribution. . Image processing : cropping each ECG image to 12 leads, process each signal of lead using gaussian, threshold_otsu, MinMaxScaler and OpenCv,Skimage in order to save it as csv file. . Create a classifier base model using ANN, CNN et RNN deep learning… Show more Processing leads signal in order to automate ECG Classifying : . Exploratory Data Analysis: extract the metadata from ECG images using EasyOCR and Keras, gather patients informations and have an idea about the ECG disease distribution. . Image processing : cropping each ECG image to 12 leads, process each signal of lead using gaussian, threshold_otsu, MinMaxScaler and OpenCv,Skimage in order to save it as csv file. . Create a classifier base model using ANN, CNN et RNN deep learning algorithms. . Hyperparameter tuning using keras tuner and optuna in order to improve model accuracy and lower model loss. Show less Processing leads signal in order to automate ECG Classifying : . Exploratory Data Analysis: extract the metadata from ECG images using EasyOCR and Keras, gather patients informations and have an idea about the ECG disease distribution. . Image processing : cropping each ECG image to 12 leads, process each signal of lead using gaussian, threshold_otsu, MinMaxScaler and OpenCv,Skimage in order to save it as csv file. . Create a classifier base model using ANN, CNN et RNN deep learning… Show more Processing leads signal in order to automate ECG Classifying : . Exploratory Data Analysis: extract the metadata from ECG images using EasyOCR and Keras, gather patients informations and have an idea about the ECG disease distribution. . Image processing : cropping each ECG image to 12 leads, process each signal of lead using gaussian, threshold_otsu, MinMaxScaler and OpenCv,Skimage in order to save it as csv file. . Create a classifier base model using ANN, CNN et RNN deep learning algorithms. . Hyperparameter tuning using keras tuner and optuna in order to improve model accuracy and lower model loss. Show less

Data Scientist

• Sep 2022 - Oct 2022

Development of a comparison and recommendation solution from e-commerce sites using the Crisp-dm methodology to make a prediction model: . Scrapping of e-commerce sites (Amazon) using Scrapy and Junglescout. . Data preparation, make the necessary transformations on quantitative variables such as price, imputation of missing data using KNN, encoding of qualitative variables in order to be consumed by our models. . Evaluate the potential of products in real time and analyze the price to… Show more Development of a comparison and recommendation solution from e-commerce sites using the Crisp-dm methodology to make a prediction model: . Scrapping of e-commerce sites (Amazon) using Scrapy and Junglescout. . Data preparation, make the necessary transformations on quantitative variables such as price, imputation of missing data using KNN, encoding of qualitative variables in order to be consumed by our models. . Evaluate the potential of products in real time and analyze the price to be optimized. . Use a recommendation model based on ratings, reviews, sales to approximate the best price or product to predict. . Evaluate the model for best accuracy value and least loss value. . Deploy the solution in a web application using Django. . Containerize the application using Docker. Show less Development of a comparison and recommendation solution from e-commerce sites using the Crisp-dm methodology to make a prediction model: . Scrapping of e-commerce sites (Amazon) using Scrapy and Junglescout. . Data preparation, make the necessary transformations on quantitative variables such as price, imputation of missing data using KNN, encoding of qualitative variables in order to be consumed by our models. . Evaluate the potential of products in real time and analyze the price to… Show more Development of a comparison and recommendation solution from e-commerce sites using the Crisp-dm methodology to make a prediction model: . Scrapping of e-commerce sites (Amazon) using Scrapy and Junglescout. . Data preparation, make the necessary transformations on quantitative variables such as price, imputation of missing data using KNN, encoding of qualitative variables in order to be consumed by our models. . Evaluate the potential of products in real time and analyze the price to be optimized. . Use a recommendation model based on ratings, reviews, sales to approximate the best price or product to predict. . Evaluate the model for best accuracy value and least loss value. . Deploy the solution in a web application using Django. . Containerize the application using Docker. Show less

Full Stack Developer

• Sep 2021 - Sep 2022

The design and development of a web application for inventory management for businesses using: . SpringBoot in Backend. . Angular in Frontend. . Docker for containerization. . Agile Scrum as a work methodology. . the use of an article recommendation system via machine learning algorithms. The design and development of a web application for inventory management for businesses using: . SpringBoot in Backend. . Angular in Frontend. . Docker for containerization. . Agile Scrum as a work methodology. . the use of an article recommendation system via machine learning algorithms.

Web Developer

• Jul 2021 - Sep 2021

Development of a web application for order management, parcel delivery and find solutions related to server failures. Development of a web application for order management, parcel delivery and find solutions related to server failures.

Database Consultant

• Jun 2020 - Aug 2020

Manage relational database and create purchase and sale invoices for paramedical products in SQL language. Manage relational database and create purchase and sale invoices for paramedical products in SQL language.