Build it so it runs on your own silicon, and so anyone can read how.
I was a musician before I was a developer, and I never fully stopped being one. Self-taught, which means every abstraction I use I had to take apart first. Curiosity is not a personality trait here; it is the method.
Melasistema is the name I ship under, not a company. Most of what I build never reaches a public repository: it belongs to the startups, agencies, companies and private clients who paid for it, and it stays behind their doors. Years of it. What is listed below is the part I am free to show — and the part I chose to write in the open.
The through-line is sovereignty. Software that runs on hardware you own, over data that never leaves it, in systems you are allowed to read all the way down. Not because the cloud is evil, but because the default should be a choice.
This directory is not the whole filesystem. The commercial work — client platforms, integrations, the systems that actually pay — is proprietary and lives in private repositories. What follows is what is mine to publish.
8 projects with a write-up, sorted by conviction rather than by stars. Stats read from the GitHub API at build time — last refreshed 2026-08-06.
This project demonstrates how you can leverage the power of Large Language Models (LLMs), specifically Gemini AI, to interact with datasets in a natural, conversational way. By combining Laravel, Gemini AI, and CoreNLP, we enable intelligent querying and dynamic responses from structured data (such as a book dataset).
MelaTheme it's build upon the Understrap framework, wich combines the power of "_s" (Underscores) starter theme with Bootstrap 5.
Luca Visciola | Resume | Full Stack Developer Voglioso di cogliere ogni possibilità per migliorare le proprie conoscenze. Tanto deciso quanto aperto nella sua visione strategica. A suo agio nella gestione di situazioni critiche.
⚠️ During the development and integration with HydePHP's static site generation process, a persistent and critical issue has been identified. ⚠️
Welcome to the Hyde Wizard repository! This magical project brings to life "The Tale of Zoe: From Doodler to Best-Seller", a whimsical journey of creativity and web development powered by HydePHP and Layouts Manager. You can find the tutorial link in the README.md!
HydePHP Events Module extends the capabilities of HydePHP, an elegant static site generator, by adding robust event management functionality. This package enables developers to easily create, paginate, and manage event pages within their HydePHP projects.
This WordPress plugin, provides a simple visually appealing and interactive timeline feature for your website. It allows you to showcase historical events, processes, or sequences in a chronological order.
Questo WordPress plugin semplifica la gestione della sezione Amministrazione Trasparente del tuo sito WordPress. Ti permette di creare e aggiornare facilmente i contenuti obbligatori per legge, assicurando la conformità al D.Lgs. 14 marzo 2013, n. 33.
Easily export your shopping cart contents as a CSV file for logged-in customers with "Simple Cart Export" plugin for Shopware 6
S-MoE — A 235B frontier model, answering from a laptop SSD. No cloud. No H100.
Seismic Mixture of Experts. A model-agnostic inference engine for fine-grained MoE models that streams experts from NVMe, so the model never has to fit in RAM.
The mountain rests cold on the SSD. Only the handful of experts each word actually needs is ever pulled into memory.
| Model | Qwen3-235B — 235B params, 128 experts/layer |
| Hardware | a standard 48 GB Apple Silicon MacBook |
| Resident RAM | ~16 GB dense trunk + a small streaming ring — not the whole model |
| Cold time-to-first-token | ~26 s, down from 92.7 s |
| Decode | 1.84 tok/s, up from 0.48 at the start of the campaign |
Point the shatter script at any compatible fine-grained MoE checkpoint and it produces
a self-describing vault: RoPE theta, routing top-k, attention geometry and lineage
all travel inside the .smoe file, so the engine reads what it is given instead of
assuming what it was built for. Swap a 235B for a 30B at runtime — same binary, same
laptop, ~13.7 tok/s.
Every byte, every struct field, every atomic flag is mapped, explained and public. No secrets. No priesthood. A door in the memory wall, and the map to walk through it.
Designed with an AI agent across many sessions: the direction came from human curiosity, the implementation depth from the collaboration.
ANDARTIS — A private fleet of tiny neural networks, trained only on your data.
The current AI landscape is obsessed with one model to rule them all. ANDARTIS rejects this.
Instead of a massive, slow, privacy-leaking generalist, ANDARTIS lets you build a private fleet of tiny, specialised neural networks trained exclusively on your own data. It is the evolution of ARKHEIN — away from general-purpose models, toward a distributed ecosystem of micro-models and conversational SLMs.
A PHP orchestrator and an Apple MLX acceleration engine talk over a zero-latency
JSON-RPC STDIO pipeline. Each Intelligence Node gets its own isolated core.sqlite —
Project A can never contaminate Project B. The Intent Blade picks the tool, the
orchestrator fetches raw rows, the Senior Analyst turns them into prose. The numbers
come from SQL, so there is nothing left for the model to hallucinate.
ARKHEIN — Command your silicon. Own your memory. A sovereign macOS agent.
Most “agents” are thin wrappers for cloud-based surveillance. They rent you your own intelligence and sell your data on the way out.
ARKHEIN is a local-first macOS agent that lives entirely on your silicon. No telemetry. No cloud gatekeepers. No data ever crosses the hardware boundary. Deep complexity hidden behind a minimalist shell.
The Vantage Hub. Knowledge is organised into sovereign silos — every authorised folder is a physical partition with its own semantic grounding, and 100% topical isolation between them.
The Mind. An eight-level cognitive pipeline — grounding, perception, context, canopy, decomposition, scratchpad, critique, generation — moves complexity out of the model weights and into the process, so small local models can follow reasoning that would otherwise need a trillion-parameter giant.
The Memory. SQLite is the single source of truth; the vector index is a disposable binary accelerator that can be rebuilt from it with zero downtime.
The Hand. Filesystem commands with surgical precision, fenced by the Silo Guard — mathematically absolute sandbox boundaries. Every action requires a verified plan. You remain the final authority.
INTENTIO — Not a chatbot. Not a cloud service. A bounded space where AI pays attention.
A local-first, private cognitive environment designed to operate inside a carefully constructed knowledge space.
It does not try to know everything. It tries to understand one context well — the one you design.
Most AI systems are built as oracles: global knowledge, generic answers, invisible data flows. INTENTIO rejects that model and treats AI instead as an interpreter of meaning inside a bounded cognitive space. Private — nothing leaves your machine. Local — models and embeddings run on your hardware. Intentional — structure matters more than scale. Inspectable — every component can be understood.
What it is not: a chatbot trained on the internet, a “chat with your docs” demo, an autonomous agent pretending to think, or a cloud-dependent service.
Private AI is not a feature. It is an architectural decision.
This is the project about attention, structure and restraint rather than spectacle — and the one that names the thread running through everything else here.
Redis Blog — Redis as the primary database, not the cache in front of it.
A modern, high-performance blog boilerplate built with Nuxt and Redis, containerised with Docker.
The premise is a small heresy: treat Redis as the primary store for a content-driven application rather than the cache you bolt in front of Postgres. Hashes for posts, sorted sets for ordering and pagination, sets for tags — the data structures do the work an ORM would otherwise do badly.
The result is a stack with almost nothing between a request and its answer. It is also the clearest example of how most of these projects start: a question about whether the default architecture is actually required, followed by a working answer.
Underwind — We love Bootstrap. We really do. So we let the wind change direction.
A WordPress starter theme built on the rock-solid _underscores foundation and
inspired by Understrap.
We love Bootstrap, we really do — but we missed the freedom and joy of Tailwind CSS. So we let the wind change direction.
Underscores gives you the correct WordPress template hierarchy and none of the opinions; Tailwind gives you the opinions you actually chose. The theme is the seam between the two, kept thin enough that you can read all of it before you start editing.
HydePHP Layouts Manager — Layouts and reusable components for HydePHP, without the ceremony.
A Composer package that simplifies how layouts and reusable components are managed inside HydePHP projects — dynamic layout switching, a component registry with default config, and theme handling that stays out of the way.
Part of a small family of HydePHP work that also includes an events module, a multilanguage experiment whose README documents exactly where it broke, and Hyde Wizard, a tutorial told as a story about a doodler who becomes a best-seller.
Archived now, but this is where the habit was formed: publish the package, publish the documentation, publish the failure too.
MelaFreelance — Time tracking built by a freelancer who needed it, for freelancers who do too.
Client, project and task management with hourly rates and payment tracking — the unglamorous software that decides whether a solo practice is actually profitable.
Built for my own studio first. That is the only reason it handles the awkward parts: mixed rates on one project, work logged days late, the difference between hours delivered and hours paid.
It is also the reference for how this site ships — the same GitHub Actions workflow shape, the same preference for a build that anyone can run locally and understand.
The larger half, and the one that is under NDA. Laravel platforms, WordPress systems, integrations, APIs and quality assurance for startups, agencies, companies and private clients. Proprietary, owned by the people who commissioned it, and absent from every listing on this page.
ARKHEIN, INTENTIO, ANDARTIS and S-MoE. Sovereign agents, bounded cognitive environments, micro-model fleets, and an inference engine that streams a 235B model off an SSD.
Redis as a primary database under Nuxt. Underwind, a Tailwind-native WordPress starter. Laravel and Gemini pointed at structured datasets.
WordPress plugins for timelines, booking and Italian public-administration transparency requirements. HydePHP layouts, events and multilanguage modules. Commerce extension work on Shopware 6.
Melasistema is not a company — it is the umbrella name for my independent work. Available for startups, agencies, and in-house teams.
Custom themes, plugins and blocks written from scratch — editorial platforms, e-commerce, membership and booking systems, and WordPress run headless as a pure content API behind Nuxt, Astro or whatever the front end happens to be. Not a page builder dragged into shape: code in a repository, versioned, reviewable, yours.
Fitness platforms, multi-tenant SaaS, REST APIs serving AI chatbots, cognitive environments for custom models, and stateless data microservices that aggregate, process and cache for everything else in the stack. Empty repository to production, including the boring parts in between.
Private RAG, small language models and on-device inference for teams that cannot or will not send their data to an API. Built on the same engines listed above, not on a vendor demo.
Architecture review, stack decisions and second opinions. Often the most useful outcome is talking a team out of building something they do not need.
Interesting problem, awkward stack, or something that should not run in someone else's cloud? Write to me.
info@melasistema.com