I have a strong motivation and passion for extracting value from data, and I love the development and engineering aspects of this work. I'm driven by the goal of delivering value through data by taking ownership of the entire process, from raw data analysis and exploration, through to production deployment and final reporting of analytics and insights, following development and DevOps best practices to ensure solutions are maintainable and scalable over time.
In my 5 years of work experience, I have developed expertise not only in data engineering but also in MLOps and software development, leading the design and implementation of end-to-end data-driven solutions that have delivered significant value and impact both in the banking sector and in product-based SaaS startup companies.
If it may be of interest or useful to anyone, below is the link to the starting repository as a starting point for developing a FastAPI backend for dubbing YouTube videos. This involves capturing and inferring voice timbre using OpenVoice and deploying it on Google Cloud Run (GCP) using Terraform, Docker, GitHub Actions, and Cloud Build
Ciao a tutti! If it can be helpful to anyone, I'm sharing a starter template repository for chatting with websites using FastAPI, Next.js, and the latest version of LangChain.
If it can be useful to anyone, I share below some code that I started to set up to build a process that allows me to extract, clean and translate texts from files
The repo comprises services for text processing and use cases for specific analyses:
Text Extraction: Extracts text from various documents.
Text Translation: Provides optional multilingual translation.
Text Cleaning: Cleans and standardizes text for analysis.
Information Extraction: Uses GPT-3.5 to extract information from texts, saving it in JSON format.
Just leaving the link to my Terraform repo here. It's set up for various GCP services that I'm using in a project. Might be useful for some of you.
VPC Configuration: A new VPC is configured with the intent to utilize all services within a single VPC.
Compute Engine Setup: A compute engine is set up to read the Cloud SQL via a private IP, configured in another module. A startup-script.sh is included, which has code to establish SSH connection directly from the local machine.
Cloud Run Creation: A Cloud Run is created in version V2 with Direct VPC Connection to Cloud SQL.
Automated Cloud Build Mechanism: A mechanism is set where the cloud build is triggered automatically following the code push to the repository that builds the image.
A cost-effective Telegram bot using Language Models like LLM, hosted on a free EC2 instance and powered by Beam Cloud's serverless REST API. This starter kit covers setup, deployment, and environment configuration, with support for the Llama 2 family of models.
Remote: Available for full-remote or hybrid work
Willing to relocate: Open to relocation
I have a strong motivation and passion for extracting value from data, and I love the development and engineering aspects of this work. I'm driven by the goal of delivering value through data by taking ownership of the entire process, from raw data analysis and exploration, through to production deployment and final reporting of analytics and insights, following development and DevOps best practices to ensure solutions are maintainable and scalable over time.
In my 5 years of work experience, I have developed expertise not only in data engineering but also in MLOps and software development, leading the design and implementation of end-to-end data-driven solutions that have delivered significant value and impact both in the banking sector and in product-based SaaS startup companies.
Résumé/CV: https://github.com/mazzasaverio (resume available upon request)
Email: saverio3107@gmail.com