I'm an industrial engineer and a data enthusiast.
I have recently finished an MSc in Data Science and AI and
I'm looking for a Data Scientist / ML Engineer position.
Hey! It's Matías Bottarini. I'm an Industrial Engineer and a Data Scientist . I've done an MSc in Data Science & AI and I'm looking for a Data Scientist / ML Engineer position.
For the past 6 years, I've been working in the Energy Sector developping software and control algorithms for energy storage systems. Recently, I added data science tools to my skillset and I'm eager to put them to work. Feel free to contact me here.
In many industries, as part of a highly performing Quality Assurance and Customer Support functions, reporting unusual events, anomalies, or process deviations requires clear, structured documentation that captures the full context: what happened, when, where, who was involved, why it occurred, and what actions were taken… This project proposes the development of an assistant generates a well-written, standardized report based on the input. Here some of the tasks carried out:
Data creation: generate a data set of reports to use them as ground truth (reference) to train the model. This was made by automating prompts through an API.
Model selection and experimentation plan : Test different Hugging Face models to evaluate its performance. Play with the generation parameters (temperature, top_k, top_p, presence_penalty, etc) and compare the results
Evaluation metrics : Text-to-text comparison, through tokenization and the attention mechanism: Bert score, bleu/rouge score, cross encoder similarity...
Model training : Using QLoRa and a SFT strategy on a small model (less than 1 billion parameters) in order to specialize it in the reporting task
Web application interface to ease accesibility and the report download
NEMS stands for Nexeya Energy Mangement System, a software created to control and manage the energy
systems of a microgrid. Particularly, it is used for managing a Hybrid Energy Storage System
consisted of Li-ion batteries and a hydrogen chain, the latter composed of a fuel cell and an electrolyzer.
Within a team of developers, I contributed to its development when I was working in Nexeya Hensoldt.
NEMS uses JS and Kibana as front-end
and C++, python and MySQL as backend.
It is deployed on Linux and uses Docker containers for hosting the Web Services.
Here a summary of the tasks that I carried out:
Code maintenance and plugin development and deployment
Time-series predictions using Random Forest and Kibana for the graph representation (Skforecast, timeLion, ...)
Development of an Optimal Control algorithm with a Model Predictive Controller
ODD is a tool used for sizing energy power systems, and more particularly, energy storage systems . Simulates for a whole year the energy balances in a microgrid which architecture is defined by the user. It is a powerful tool for the commercial and back office services. Here some of its features:
Downloads real meteorological data in order to estimate the renewable energy production (PV and Wind Turbines)
Calculates the main KPIs to give a global vision of the system's performance and system's costs
Automatically finds the best size thanks to a PSO algorithm that minimizes the CAPEx and OPEx costs
Developed in PyQt5 as a desktop software, has a friendly and responsive Graphical User Interface (GUI)
