Stars
RAG workflow-friendly enhancement of Marker that converts PDFs into a single-file markdown equivalents. It handles GPU and PyTorch configuration, document splitting and chunking, image BASE64 embed…
Constructs a stylometric profile from an author's writing corpus, then applies that profile to rewrite or generate new text in the same voice. This enables an LLM to apply a specified writing style…
CityLife is an isometric city builder that functions as a life-planning simulator - place “activities” as buildings, connect them by roads, and explore how distance-weighted relationships shape inc…
Lightweight Streamlit-based log time-series visualizer that pulls data directly from Ubuntu journalctl, chunking it by user-defined intervals and displaying intuitive heat maps to highlight changes…
Excel-DNA add-in that turns Excel into a transparent, governed workspace for preparing RAG-ready data (ingestion, chunking, metadata, QA, and export). Still in the early stages but benefiting from …
A metadata-based and LLM-augmented publication renaming toolkit. Automated metadata extraction and optional LLM ingestion of doc content makes up for poor or missing metadata. Useful in RAG pipelin…
Machine learning and deep learning, natively inside Excel — built to support our clients in learning about ML through experimentation. Contains very valuable and proven-to-work ExcelDNA patterns th…
AI-powered threat monitoring solution for small/SME networks, leveraging Snort, ntopng, pfSense, pfBlockerNG, and LLM analyses to detect, assess and take action in real time against malicious actor…
This repository provides a suite of high-performance, thread-safe Excel User Defined Functions (UDFs) developed using the Excel-DNA framework in C#. These functions significantly extend Excel's bui…
Linux CLI tools to compare text files and find nearest neighbours across large directories using TF‑IDF or SimHash, with optional dedup workflows, useful in RAG pipelines to remove duplicate docume…
Find similar text files in a repository and sort by similarity
Small and simple directory synchronizer (a BASH script)
This repo is a z88dk C library for the NABU PC (personal computer). It includes efficient C functions for video, audio, hcca, keyboard, and more for the NABU Computer.