I am a Python Developer and independent AI Researcher, specializing in AI Engineering, NLP, and Deep Learning, with a strong foundation in Sound Engineering and Technical Sound Design.
Initially, I pursued studies in Computer Science and Archaeology, but my curiosity and passion for technology led me to specialize in audio and DSP algorithms. Over three years, I focused on Technical Sound Design, Sound Engineering, and Composition. This led me to develop advanced DSP algorithms, spatial audio, and AI-driven sound systems for video games using C++, JUCE, and Unreal Engine.
Later, I returned to Computer Science, obtaining certifications through CodeAcademy, reinforcing and expanding my expertise in AI and machine learning.
However, my early studies in archaeology and ancient cultures—Greek philosophy, Egyptian architecture—never really left me. I’ve continued to weave them into my growing curiosity for physics, philosophy, and the fundamental nature of reality, which often drives me to dig deeper into original sources—academic papers, written articles, primary research—and to connect ideas across different perspectives.
This multidisciplinary lens lets me approach AI not just as a tool, but as a reflection of deeper human questions—merging cognition, history, and theoretical sciences to explore what AI can truly become.
- Working on freelance projects as Prompt Engineer for Outlier.ai and Mindtel-Innodata, where I still freelance, my most important project is Mistral Le Chat (Italian version) – Contributed to DocsGPT and the NLTK library.
- Developed AI-driven solutions for model optimization with RLHF.
- Built NLP chatbots
- AI App powered e App Audio.
- Designed advanced spatial audio and DSP algorithms for video games. Some of these algorithms were used as the basis for Spatial Audio for the unDAW project Screenshot
- Sound Designer for Cookie Cutter and other indie titles on Itch.io and Steam. My collaborative experiences include participation in Game Jams and voluntary projects, notably a collaboration with fellow freelancers. We established a Discord group, Remote Access Games, which, despite being discontinued due to divergent member perspectives, serves as a significant portfolio piece.
- I've written music for bands and for personal projects, released on platforms like Bandcamp.
- I conducted audio research for my former academy, Mas Academy, which was published on YouTube and used in real-world projects and internships.
- Founder of BG Art House Studio, an indie game company. (I stepped down from the company I founded shortly after.)
- Collaborated as a Beta Tester for Ableton Live 12, providing feedback on the integrated spatial audio, similar to that of Nuendo 12, through the beta testing platform.
- Worked as a freelance journalist, publishing (two articles) in the video game industry.
- Freelance developer, working on contract or occasional projects.
- I am part of a community of independent AI researchers, granting me access to a wealth of AI-related resources beyond my personal research.
My research focuses on AI memory structures, exploring how neural networks can develop long-term memory systems inspired by human cognition. I’ve published independent studies on Zenodo, proposing models that integrate working memory, short-term, and long-term memory into AI architectures.
I have also released two major AI research projects on Ready Tensor Inc:
- NoVa – A deep dive into ChatGPT-like models, analyzing their internal mechanisms and proposing structural improvements.
- Omnia – An auditory neural network, extending AI cognition into sound-based perception, complementing my research on memory and sensory-driven AI learning.
- You can find further publications here.
These are some of my Scientific essay on Medium exploring the world of Deep Learning, but also expanding into broader themes such as philosophy, quantum physics, love, and more.
I am currently exploring generative AI and deep learning, advancing my work in AI cognition, neural architectures, and memory-inspired learning, aiming to bridge intelligence, creativity, and human-like memory systems through research, development, and open-source contributions.
- Prompt Engineering Lab, an interactive sandbox on Hugging Face Spaces, designed to test and refine prompts.
- See my Badges
- These are all my certifications (some ongoing), acquired throughout my career, in addition to my incomplete university studies. The links are accessible with an active LinkedIn profile.
- Python, C++
- Django, Flask
- JUCE (C++ Audio Framework)
- TensorFlow, Keras, SmolAgents (Hugging Face)
- Stable-Baselines3
- Natural Language Processing (NLP): NLTK, Hugging Face Transformers
- Deep Learning Architectures: LSTMs, CNNs, Transformer, Reinforcement Learning (RL), DRL, RLHF
- Reinforcement Learning Frameworks: Stable-Baselines3
- AI Memory Systems: Elastic Weight Consolidation (EWC), Continual Learning , Transformers with Self-Attention
- AI Agents: SmolAgents
- Prompt Engineering & Large Language Models (LLMs)
- Generative AI & Multi-Modal AI
- PostgreSQL, SQLite
- Docker
- CI/CD Pipelines, GitHub Actions
- Digital Signal Processing (DSP)
- Spatial Audio & 3D Sound Design
- Game Audio Implementation (Wwise, Unreal Engine)
- Music Composition & Sound Synthesis
- DAW (Ableton Live, Steinberg Nuendo, Izotope RX)
- Docker, Linux Administration
- Git, GitHub, Shell Scripting
- Jupyter Notebook, Colab
- Remote Jobs App: Explore companies offering remote jobs (Flask, SQLAlchemy).
- MovieRanker: Filter and rank movies using TMDB API (Django, TMDB).