Connects to your Trello account to fetch cards from your 'Inbox' list. Uses locally run AI models (embeddings and LLMs) to suggest appropriate destination lists for classification, helping you organize tasks efficiently. Also includes calendar, upcoming task views, checklist management, and a 3D mind map visualization.

An interactive Natural Language Processing suite for exploring, analyzing, and visualizing text data. Features include sentiment analysis, entity extraction, summarization, keyword mining, and rich result visualization—all powered by local and open-source models. This is in Beta and no all servers are running. This is currently CPU bound if you would like to see a gpu version let me know.

Uses local language models to discover knowledge hidden and forgotten in your Trello boards, cards and lists. Extracts insights and connections from your existing data.------------ This is in beta and currenly only loads one gguf , this is using a recompiled open cl but is limited to 8gb vram . next revision will have the top fastest gguf llm loaded and able to be switched between. currently changing the gguf is disabled for repair in next version.

A focused, silent reading tool built on RSVP (Rapid Serial Visual Presentation) principles. Mindstride helps you silence your inner monologue and read faster—one word (or chunk) at a time.

Upload screenshots to extract text using OCR (Optical Character Recognition) and then generate a concise summary of the extracted text. Supports various image formats. Results are streamed back to the page.

Frustrated by Apple Weather and AccuWeather getting it wrong, I set out to rebuild the Dark Sky experience with truly hyperlocal forecasts—this alpha pulls from multiple forecast APIs now, and soon will overlay DOT-mounted sensors and community PWS feeds. Next up: ML-based smoothing and predictive models running on local LLMs accelerated by Intel Battlematrix GPUs.

RustBerry is a procedural resource gathering simulation, where Python-based agents build villages, gather resources, and adapt to a dynamic world. This companion tool demonstrates the generative art pipeline—creating trees, village layouts, and asset previews—designed for RustBerry’s resource engine. The generator produces a seed file format that RustBerry understands, enabling seamless integration of procedurally generated assets into the game world. The most developed part of this tool is the top-down tree generation, offering diverse, randomized tree structures for use in the simulation. Once the core simulation is running, I’ll stream a live video of the LLM models’ logic, choices, and visual output—so we can watch the game world evolve in real time as the AI makes decisions.

A visual context, antonym, and synonym data mining explorer. Provides comprehensive knowledge management and linking features for discovering hidden connections in your data.

A lightweight version that only creates a small tree of connections, while the main version can create such large patterns it can consume too many CPU and memory resources.

A 3D implementation of Conway's cellular automaton visualized on a sphere. Explore patterns and how they evolve over time in a dynamic WebGL environment.

A WebGL implementation of the classic 3D Pipes screensaver animation. Watch intricate pipe systems grow in real-time.----- my goal is to hook its random and sprite input to something organic allowing for a growth pattern to be driven by an external var force. . currently experimenting using telescope and sun luminace as seed data.

A collection of tools focused on Natural Language Processing powered by Large Language Models. Includes capabilities like entity recognition, sentiment analysis, text summarization, and more for text-based data.