ABXK.AI Projects

Applied AI research systems designed, built, and operated by ABXK.AI.

AI Prediction

These projects use machine learning to study pattern formation and predictive behavior in data under real-world conditions. We study how AI models perform when they face real-world problems like noisy data, changing conditions, and missing information.

Internal

AI Trading Platform (Research)

A research platform that studies how AI models behave with financial time-series data.

We focus on how models perform, how to validate them properly, and where they fail under real market conditions.

This is for analysis only — we don't provide trading signals or promises. Any trades are done internally for research and to watch how the system behaves.

(Not a live trading product or external service.)

  • ✓ Neural network models for sequential market data (LSTM, Transformer)
  • ✓ Multi-asset datasets across diverse market symbols
  • ✓ Walk-forward validation and robustness testing
Learn More
Internal

AI Racing Data Analytics (Research)

A research project studying how AI models work with telemetry data from physical systems like racing cars.

We look at uncertainty, model limits, and trade-offs in evaluation — not at making faster lap times.

All outputs stay internal. This project doesn't give advice or decisions for anyone outside our team. It's a case study in methodology.

  • ✓ Telemetry-style time-series analysis
  • ✓ Model behavior under noisy and constrained inputs
  • ✓ Evaluation limits in physical-system contexts

This page documents applied AI research systems designed, built, and operated by ABXK.AI. The projects listed here are not products, tools, or tutorials. They are internal systems created to explore how AI behaves in practical, real-world settings under constraints such as data quality, cost, reliability, and long-term operation.

Each project applies AI within a specific context—such as data-driven research or experimental modeling—with an emphasis on system design, evaluation, and observed behavior, rather than on tools, trends, or theoretical results.

Some projects are active, while others are experimental or research-focused. All share the same purpose: to test assumptions, understand trade-offs, and document what works—and what fails—when AI is applied in real-world workflows.