Data Science Portfolio

I’m always practicing and learning new tools to improve my data scientist and coding skills. In this page you can find some exercises I’ve done in Cognizant as a data scientist, in the Ironhack Data Analytics bootcamp and on my own. Feel free to have a look at my personal GitHub account as well!

Also, I’ve deployed some demos in Streamlit with data science models I’ve developed. Click here to run them online.

Invited talk about GenAI at FOSSCILT 24:

I was an invited speaker at the FOSSCILT 24 conference:

https://fosscilt.icfoss.org/

We gave a presentation about the performance of LLMs with mathematical operations. Click here to watch it it.

Azure Percept object detection case study

I’m collaborating in many Computer Vision case studies at Cognizant. As a data science consultant, I’m in charge of creating deep learning models or other AI approaches to solve customer needs.

Microsoft offered a bootcamp to test Azure Percept, an IoT device for creating edge AI solutions. The case study we’ve created is a general fruit detection model using Azure Percept, Custom vision and IoT Hub. In this video you can have a look at our approach.

In the Microsoft’s Internet of Things Blog we have published an entry describing our approach. Click here to read it.

Automated maintenance - Rust detection

Using deep learning approaches, I’ve contributed to the automatic rust detection in a case study. Below I show blog posts and the description of the solution as shown in the Cognizant media:

Ironhack bootcamp projects

You may click here to see the projects I’ve created for the Ironhack Data Analytics bootcamp and a brief description. Below I just list the GitHub repo links:

  1. Hangman game, programmed in Python (moddable)
  2. A word frequency based analysis of sharks attacks
  3. 500 best albums of all time / Spotify API project
  4. GeoSpatial Data Project / Choosing the best location for a gaming company
  5. API chat project and text sentiment analysis
  6. Kaggle competition - Predicting diamonds prices
  7. FINAL PROJECT: Dice Scores Recognition in image/video

Self-paced projects

Also, I highlight the following self paced exercises I’ve done for practicing and learning purposes that are public:

Feel free to have a look at other repositories here with other exercises, learning notebooks & code or just simple tests, even if they are uncompleted or not interesting enough to be shown in the profile.