f97.be!

I have been writing software since I was 16 years old. My foundation was built in the era of 8-bit microcomputers, cutting my teeth on the Commodore 64 and ZX Spectrum. From those early systems, I expanded into the world of heavy iron, navigating mainframes like the PDP-11 and working through the evolution of the Unix ecosystem that shapes so much of modern computing.

As the hardware scaled, my tooling evolved right alongside it. I progressed organically through the foundational logic of Basic, the bare-metal precision of Assembler, the systems-level control of C and C++, and the robust object-oriented architecture of C#.

Today, my core focus is the modern Apple ecosystem. Since the early 2020s, Swift has been my primary language. Apple succeeded in creating a toolset that empowers an independent developer to seamlessly engineer highly performant, enterprise-grade applications. Swift perfectly bridges the gap, offering the sheer power required for complex architecture with the modern syntax needed to build efficiently. Served with a healthy side of Python for dessert when it's time for analytical heavy lifting, AI integration, and scripting.

I leverage this stack to design and architect native macOS software from the ground up, prioritizing fluid user experiences, local performance, and resilient design.

Below is a collection of my recent applications.

NeuralWeave AI is a native, local-first macOS application for intelligently organizing, searching, and curating large image archives. It combines professional archive management with private, on-device artificial intelligence, allowing users to find photographs not only through filenames, tags, dates, locations, and metadata, but also through their actual visual content. Natural-language descriptions in German, Dutch, or English can be used to search the entire library, while visual similarity search identifies related photographs, alternative versions, and near-duplicates.

The application imports common image formats together with EXIF and GPS metadata, performs OCR, generates high-quality thumbnails, and organizes images into collections and albums. It also provides advanced filtering, map-based browsing, batch metadata editing, duplicate cleanup, export tools, backups, and a full-resolution image viewer. Its AI features use SigLIP2 and Apple MLX directly on the Mac, while image records and vector embeddings are stored locally in an SQLite database. NeuralWeave AI can therefore be described as a private, AI-powered digital asset management and semantic image search system designed for extensive photographic and archival collections.

f3 is a native solar-energy forecasting system for macOS that combines machine learning, weather data, physical irradiance modelling, and live inverter measurements. It blends forecasts from Apple Create ML with a physics-based GTI model to produce calibrated photovoltaic yield predictions and realistic uncertainty ranges. A background agent continuously collects production data, retrains and validates the models chronologically, evaluates forecast accuracy, and publishes an interactive bilingual dashboard with current output, seven-day forecasts, historical performance, and transparent methodology. Built entirely in Swift, f3 provides a scientifically grounded and fully automated view of how much solar energy a photovoltaic installation is likely to generate.

ScriptAIX is a multi-tenant, AI-powered transcription and document analysis platform engineered specifically to process, transcribe, structure, and analyze handwritten historical manuscripts and archival records. Designed as an independent, comprehensive pipeline, it bridges raw digitized scans and modern archival data formats, providing an end-to-end workflow from image ingestion to standardized XML and knowledge graph generation.

Register here for the free Beta.

VoluminaX bridges the gap between personal library management and digital publishing. It is a dedicated desktop application built for meticulous book and library cataloging. Beyond simply organizing a collection locally, VoluminaX functions as a static website engine, allowing you to generate and publish comprehensive, web-ready catalogs directly from your local database.

Engineered for a frictionless coding and writing experience, EEE is a native macOS text and code editor suite. It integrates a fast, dual-pane file manager directly into the workflow. Built for modern development, it includes deep Language Server Protocol (LSP) integration for intelligent coding assistance, built-in LaTeX tools, and a zero-dependency SQLite database inspector, allowing you to manage code, files, and data from a single unified interface.

Willy Willy brings frictionless audio streaming directly to the macOS status bar. Originally designed as a lightweight, native player for Belgian radio stations like Willy Radio, Qmusic, and JOE, it stays out of your way while you work. It also features seamless Sonos protocol compatibility, allowing you to control multi-room audio playback directly from your Mac's menu bar without needing to open a heavy standalone controller app.

Don't have a Sonos, I've got you covered. Turn any computer into a network player with Willy Willy Python Player.

fm is a powerful native file manager for macOS that combines flexible multi-pane navigation with fast, precise, and secure file operations. Designed for professionals and power users, it offers tabbed browsing, background transfers, advanced search, folder synchronization, batch renaming, archive tools, integrated terminals, Git support, remote storage access, and comprehensive previews. Tamper-evident audit logs and cryptographically sealed File Vaults provide an additional layer of traceability and data integrity—all within a responsive interface built specifically for the Mac.

fmmav is a read-only integrity and audit verification application for macOS. It independently examines the cryptographic audit logs and File Vaults created by fm, validating SHA-256 hash chains, digital signatures, file seals, manifests, and mirrored copies. Changed, missing, replaced, or touched files are clearly identified, while detailed findings and human-readable reports make every verification result transparent. fmav provides reliable evidence that archived data remains complete, authentic, and unchanged.

GeneaFlow is a native genealogy application for macOS that transforms Gramps and GEDCOM data into clear, interactive family trees. It lets researchers explore ancestors, descendants, and complete family lines; search and edit people and relationships; examine genealogical statistics; and detect chronological, biological, and structural inconsistencies. Its integrity-aware editing system preserves data outside the fields being changed, creates backups, and verifies every update before writing it to the original Gramps file. Finished family trees can be printed, exported as vector PDFs, or published as privacy-protected interactive websites, making GeneaFlow a complete environment for exploring, improving, preserving, and sharing family history.

The applications presented here - NeuralWeave AI, fm, fmav, f3, and GeneaFlow - are my release-ready projects. Each has progressed beyond the experimental stage into a complete macOS application designed for practical, dependable use. They represent years of work across artificial intelligence, file management, data integrity, forecasting, and genealogy.

These releases are only part of a much broader software portfolio. Several additional applications are already in development, including eCMS, TranslAItor, and Trueframe. Each explores a different set of ideas and challenges, but they all follow the same principles: purposeful design, technical depth, privacy, and software that gives its users genuine control.

As an independent developer, I continue working on these projects whenever time permits. Some are already well advanced, while others are still evolving as I refine their concepts, architecture, and interfaces. They will be added to this page when they reach the level of maturity and reliability required for release. There is much more in the pipeline.