Hi, I’m Gabriele.

I’m the founder and lead A.I. and M.L. engineer of G.M.S.C. Consulting, an A.I consulting firm whose mission is to enable small and medium software companies to ride the unavoidable A.I. revolution, rather than drown in it.

This page will tell you more about who I am, what I did, and what I can offer you and your company.

Here’s a brief portfolio of Machine Learning projects I worked on

  • We used computer vision and NLP to digitalize old Italian ID cards with challenging watermarks and open sets values

  • Development of biometric based access-control software. Way before this kind of technology was almost a commodity

  • Development of algorithms for estimating an optimized trading strategy for bonds

  • Development of a system that by leveraging smart OCRS extracts relevant information from unstructured documents to be feed into a robot process automation system

  • I wrote from scratch a render engine to create synthetic data, which replaced real but hard to collect data.

  • I created from scratch a computational framework designed to train and run efficiently machine learning models applied to IOT data.

Here at G.M.S.C. Consulting, we will help you ride with the wave, in this case, the unavoidable AI revolution, rather than drown in it.

My Recipe to Delivering Extraordinary Value to Our Customers

After 3 STEM degrees and having completed many successful projects as lead machine learning and A.I. engineer, I found my 6 points recipe that delivers extraordinary value to my customers.

  • AI solutions that deliver accurate and robust performance under expected boundary conditions and can detected when the input data is outside their working domain.

  • Writing software that complies with best practices and sound software architecture makes a huge difference once the AI model is in production. Less bug fixing and more uptime.

  • The simplest A.I. solution is the best A.I. solution because it’s more stable, explainable and faster.

  • Instead of what they want. The best medicine for your problem usually is not exactly what you asked.

  • Software and A.I. must be built together in a way that is efficient, scalable and fast

  • I take pride in delivering good software, that can be easily understood and maintained. This entails documenting the code and adding extensive test suites for early bug detection and reliable software.

Want to work with us?

Check out our packages
Get started

Did you say languages?

Here’s a list of programming and natural languages that I use

Programming languages

  • Python: Master

  • C++: Expert

  • Matlab: Expert

  • .Net: good

Operative Systems

  • Debian (Ubuntu): Proficiency with both GUI and Terminal

  • Windows: Good

Natural Languages

  • Italian: Native Speaker

  • English: Proficient Speaker

  • Chinese Mandarin: Traveling and basic work interactions

  • Spanish: Traveling and basic work interactions

Let’s talk about Tech Stack

Python is my language of choice for all AI related projects.

Python provides a vast amount of ML libraries that make any ML project much easier to build.

A few major examples are Spacy (a library for Natural Language Processing), H2O (Auto Ml), Tensorflow (Low level Neural Networks library), Scikit-Learn (a collection of traditional and novel machine learning models), SciPy (Signal processing), Nltk (Natural language Processing), Pandas (Data Science)

Deployment & DevOps

The deployment of ours ML solutions mostly fall into 3 categories:

  1. Cloud computing.  The ML solution is packaged in a docker container and exposed to the other cloud service by RestApi or Message Processing. DevOps are managed by Kubernetes + CICD (ex: Jenkins or code pipeline) + git

  1. Local computing - separate program.   The ML solution is once again packaged in a docker container which interacts with other running applications by Socket and IPC. DevOps are managed by CICD (ex: jenkins) + git

  2. Local computing - same program.   The ML solution is wrapped by a C or C++ layer thus becoming a C library that can be embedded in almost any application. DevOps are managed by CICD (ex: jenkins) + git

Learn more

Book a call with me.

Don’t let technology overwhelm you. Let us help you transform your business and make it thrive.