Matthew Charles Vladislav Busel

tokio-prompt-orchestrator. Multi-core, Tokio-native orchestration for LLM pipelines.

Every-Other-Token. A real-time LLM stream interceptor for token-level interaction research

LLM-Hallucination-Detection-Script. A comprehensive toolkit for detecting potential hallucinations in LLM responses. Compatible with any LLM API (OpenAI, Anthropic, local models, etc.)

fin-primitives. Financial market primitives — price types, order book, OHLCV, indicators, position ledger, risk monitor

FinRL_DeepSeek_Crypto_Trading. This repository contains our solution to Task 1 of the FinRL Contest 2025: developing a high-performance crypto trading agent using reinforcement learning and LLM-derived market signals.

Special-Relativity-in-Financial-Modeling. The first end-to-end C++20 implementation of special-relativistic geometry applied to financial OHLCV data. Computes Lorentz factors, spacetime intervals, Christoffel symbols, and geodesic deviation signals from live market data.

Reddit-Options-Trader-ROT-. A modular research pipeline that turns trending Reddit discussions into structured market events and options trade ideas.

llm-cpp. The C++ LLM toolkit. 26 single-header libraries for streaming, caching, cost estimation, retry, and structured output. Drop in what you need.

Token-Visualizer. The ultimate tool for analyzing, visualizing, and optimizing your LLM prompts

fin-stream. Real-time market data streaming primitives — 100K+ ticks/second ingestion pipeline

Quantum-Neural-Network-Model-The-Future-of-AI-Cognition. Quantum computing is revolutionizing the way we think about processing power. Now, imagine applying the principles of quantum mechanics to a neural network—one that mimics and accelerates human cognition, surpassing current deep learning capabilities

Mycelium-Based-AI-Integration. This repository explores the intersection of synthetic biology and artificial intelligence by leveraging mycelium as a bio-computational substrate. The project aims to give AI systems a "body" by integrating machine learning models with the natural properties of mycelium, such as signal conduction, adaptive growth, and environmental reactivity.

rust-crates. Production Rust libraries for AI agents and financial systems — zero panics, typed errors, aerospace-grade test coverage

agent-runtime. Unified Tokio agent runtime — orchestration, memory, knowledge graph, and ReAct loop in one crate

tft-synapse. A lightweight TFT decision engine that scores Augments, Items, and Board choices using rule-based heuristics and human-readable YAML meta files.

llm_affector. An async Rust library for LLM-based content analysis, providing hallucination detection and code critique functionality. Built with Tokio for high-performance concurrent operations.

ROT-TECH-PDF.

Secondbrain. This Obsidian vault is a curated collection of notes, resources, and projects related to Machine Learning, Artificial Intelligence, and Cloud Computing. It serves as a personal knowledge base to organize and access information efficiently.

LLMTokenStreamQuantEngine. A low-latency, C++-based simulation engine that ingests token streams from an LLM in real-time, maps semantic token meaning to trade signals, and triggers micro-adjustments to a trading algorithm on a fractional-time (sub-second) scale.

-Event-Horizon-Siphon. The Event Horizon Siphon is a theoretical framework for a compact, ultra-dense energy engine that harvests relativistic and quantum energy outputs generated in the moments prior to matter crossing a black hole's event horizon. This engine leverages a controlled interaction between a miniature black hole and neutron star matter to extract otherwise

Context-First-AGI-Cognitive-Context-Engine-CCE-Prototype. The Context-First AGI project is a radical departure from traditional LLM-based models. Rather than focusing on generating language through token prediction, we’re developing a cognitive model that builds thought before output. Our goal is to create an AGI that mimics human cognition.

Project_Perfume. Fragrance Chatbot with Email Functionality This project is a Python script that allows you to interact with a fragrance expert AI and get recommendations on various fragrances. It also provides the option to email the responses to yourself or others. Features Chatbot Interaction: Ask questions about fragrances, such as popular choices.

HelixRouter-adaptive-async-compute-router-. HelixRouter: an adaptive Tokio job router that switches execution strategies based on size vs scaling potential.

llm-stream. Zero-dependency single-header C++ library for streaming OpenAI & Anthropic LLM responses. Drop in llm_stream.hpp and go.

srfm-python. Python SDK for Special Relativity Financial Modeling — Lorentz factors, spacetime intervals, geodesic signals on OHLCV data

llm-sync. CRDT and vector clock primitives for distributed LLM agent state synchronization

tokio-memory. Agent memory primitives — episodic, semantic, working memory with decay and consolidation

tokio-agent-memory. Tokio-native agent memory — episodic, semantic, working memory with decay, persistence, and multi-agent bus

llm-budget. Autonomous cost governance primitives — hard budget enforcement across agent fleets

-V.I.B.E.-Vibrational-Inference-for-Biological-Events. VIBE is a speculative research concept exploring the use of ultra-sensitive acoustic sensors and AI/ML to detect cellular mutation events, such as early-stage cancer, based on the minute vibrational signatures they produce.

adaptive-task-scheduler. A proof-of-concept runtime for dependency-aware, forced task promotion in Rust using `tokio`.

spacetime-storage.

tokenviz. TokenViz — A CLI tool to visualize token usage in OpenAI prompts, helping developers optimize and understand prompt structure for better model performance.

Raven-Intelligence-Model-v1-.

DreamML. Dream-Based AI Model for Creative Problem Solving, Emotional Insights, and Unfiltered Thought Generation

NeuroPulseML-AI-for-Human-Electrical-Signal-Diagnostics. NeuroPulseML is an experimental machine learning framework designed to detect early-stage medical conditions by analyzing the electric signals emitted from the human body.

Project-JarvisBox. JarvisBox** is a voice-activated, AI-powered, hardware-integrated assistant designed to replicate the functionality of a real-world "JARVIS" system. It combines **LLM intelligence**, **multimodal input/output**, and **physical expressiveness** into a standalone smart terminal.

-real-time-neural-pattern-interpretation-and-orchestration.. This MVP lays the groundwork for a living intelligence model, where AI doesn’t just process data, it remembers, learns, and aligns with human values in real-time.

Galactic-Resonance-Network-GRN-. Cosmic-scale communication & mapping through frequency-based acoustic beacons

Projects. Weather-Based Camera Settings Recommender This Python script fetches weather data for a specified city using the OpenWeatherMap API and then suggests optimal camera settings based on the current weather conditions. It also sends an email containing the weather information and camera setting recommendations to a specified recipient.

Probabilistic-Future-Event-Modeling. This project explores a novel approach to forecasting by leveraging probabilistic reasoning to model future events before they occur.

Project-SEC-Filing-Analyzer-Tool. Financial Data Anomaly Detection and Forecasting Project Overview This project involves the development of a machine learning model to analyze financial data, detect anomalies, and forecast future trends. Using a variety of models from scikit-learn, such as RandomForestRegressor, KMeans, and IsolationForest.

awesome-llm-infra. Curated list of production LLM infrastructure — runtimes, clients, observability, safety, memory, and finance

llm-cost-dashboard. Real-time terminal dashboard for LLM token spend — cost/request, projected monthly bills, per-model breakdown (ratatui)

llm-wasm. LLM inference primitives for WebAssembly — cache, retry, routing, guards, cost tracking, templates

llm-bench. Universal LLM provider benchmark CLI — compare OpenAI and Anthropic on latency, cost, quality, and throughput

mem-graph. Knowledge graph primitives — entities, typed relationships, properties, BFS/DFS traversal

prompt-observatory. Unified LLM interpretability dashboard — real-time token streams, hallucination scoring, and prompt cost analysis

llm-parse. Zero-dependency single-header C++ response parser for LLM APIs. Extract structured data, markdown, code blocks, and more from model output.

wasm-agent. ReAct agent loop for WASM/edge — Thought-Action-Observation with tool dispatch

llm-json. Zero-dependency single-header C++ library for extracting and validating JSON from LLM outputs. Handles malformed responses gracefully.

rot-signals-api. REST + WebSocket API for Reddit Options Trader signals — real-time for paid tiers, 15-min delayed free tier

llm-diff. Output diffing and versioning for LLM outputs — semantic diff, version store, lineage tracking

llm-rank. Zero-dependency single-header C++ reranking library for LLM-based result scoring and relevance ranking.

tokio-llm. Tokio-native LLM client — OpenAI + Anthropic with retry, circuit breaker, streaming, and budget enforcement

llm_affector.. Rust

Every-Other-Token.. Rust

Receiver. Codebase for the Lora Receiver board

Mattbusel. Config files for my GitHub profile.

adaptive-task-scheduler..