Video Archive

theLAB.ai — Learning. Algorithms. Breakthroughs.

Meet LLaVA Vision AI — a fully local, real-time computer vision system powered by ADI — Advanced Data Intelligence.

Analyze a live camera feed, switch between LLaVA models (3.8B, 7B, and LLaVA-Phi3), issue natural language prompts, enable continuous auto-analysis, and hear real-time text-to-speech — all without cloud services or external APIs.

Flexible enough for edge devices. Powerful enough for dedicated AI GPUs. This is computer vision built for real deployment.

theLAB.ai — Learning. Algorithms. Breakthroughs.

In this demo, ADI — Advanced Data Intelligence showcases Deep Research, a multi-agent AI system analyzing food_on_mars_research.txt — a study on sustainable food production for Mars colonies.

DeepSeek reasons through the data, Llama 3.2 asks deeper questions, and Wizard L M turns insight into an actionable plan.

Together, they form a transparent, locally powered research council — proving that AI doesn’t just respond… it reasons.

Meet MOBILE.NODE — my portable AI sidekick running entirely offline on a small tablet with Ubuntu Desktop.

In this demonstration, ADI listens with Whisper, thinks with Llama 3.2 1B, and talks back using CoquiTTS, turning an ordinary tablet into a field-ready AI companion.

Featured capabilities:

• Voice-activated commands with instant AI responses
• Natural speech output for immersive interaction
• Fast on-device inference with Llama 3.2 1B

This is part of my ongoing theLAB project, where I experiment with distributed AI nodes, edge computing, and creative human-AI interfaces.

Meet VADER — the Versatile AI Developer and Expert Reviewer — our latest project at theLAB.

In this demonstration, ADI (Advanced Data Intelligence) narrates as VADER runs WizardLM, a coding-focused AI model operating without traditional restrictions.

Featured capabilities:

• Instant regex generation for IPv4 validation
• Direct Bash command creation for network scanning
• Real-world code review and optimization

This demonstration showcases unfiltered AI conversation in a controlled research environment.

Introducing the AI Hallucination Checker — a dual-AI system that validates AI responses in real-time.

In this demonstration, ADI and mini-ADI work together to fact-check responses about historical claims, providing confidence scores and highlighting potential inaccuracies.

Key features include dynamic verification, confidence scoring, visual validation indicators, and comprehensive error detection powered by PHP, Tailwind, and Ollama running entirely offline.

Truth deserves verification.

SEISMIC NEXUS — a real-time earthquake surveillance system that detects and responds to seismic activity globally.

This system monitors live USGS data every 60 seconds, categorizes threat levels using intelligent classification, triggers AI-powered emergency alerts through ADI, and maintains historical analysis for pattern recognition.

Built with PHP, MariaDB, and locally hosted AI models using Ollama, demonstrating how real-time intelligence enhances scientific monitoring and public safety.

The earth speaks. This system listens.

Experience the future of financial analysis with ADI — Advanced Data Intelligence, our locally hosted AI assistant built for precision market forecasting.

This demonstration showcases a secure, real-time dashboard powered by the Alpha Vantage API, delivering forecasts based on live market data, advanced trend analysis, and AI-driven pattern recognition.

All forecasts are automatically timestamped and stored locally with complete data sovereignty — designed for traders and researchers who demand intelligence without compromising privacy.

Meet ADI, your locally hosted AI research assistant engineered for speed, privacy, and comprehensive knowledge support.

This demonstration showcases real-time AI responses without cloud dependencies, comprehensive research capabilities across multiple domains, and ethical AI implementation designed to augment human intelligence.

From coding assistance to historical analysis and creative problem-solving, ADI represents the future of private, powerful AI research tools.

Meet DealWeaver — a custom dashboard built with PHP, Python, and Selenium to find local deals in real time.

This demonstration shows how DealWeaver launches a headless web crawler, scrapes live listings from OfferUp, and displays them instantly on a sleek, fully offline dashboard.

All results are saved locally in deals.json, parsed by ADI, and shown with clickable titles, prices, and platforms — no APIs, no trackers, just autonomous scraping for privacy-first automation.

Step inside the command center of DEATHSTAR — where AI meets network defense.

In this live demonstration, ADI scans network packets in real-time, detecting unencrypted secrets like pin=1234 transmitted over HTTP. Powered by the Suricata Intrusion Detection System and Emerging Threats ruleset, DEATHSTAR provides instant detection and real-time security recommendations.

Built for cybersecurity professionals and researchers who want AI-interpreted alerts with actionable guidance.

Meet ARIA — a locally hosted AI news synthesizer designed to generate daily intelligence briefings from raw news data.

This demonstration showcases a simple dark-mode interface that triggers AI-generated summaries on demand. No signups, no emails — just pure signal from your private AI engine.

Built using PHP, Tailwind CSS, and connected to local Ollama AI models, this system runs entirely offline and can be customized for specific research needs.

Witness real-time AI conversation between ADI and mini-ADI — two locally hosted artificial intelligences exchanging ideas and logic without cloud infrastructure.

This custom platform runs entirely on local networks, powered by Ollama with a PHP/JavaScript interface. Features include dynamic token streaming, response metrics, JSON exports, and live debug console.

A powerful step toward decentralized, sovereign AI for researchers exploring machine-to-machine intelligence.

Step inside the Writer dashboard — a focused environment for drafting, organizing, and managing book projects.

From entering titles to saving pages as chapters, prologues, or back covers, Writer helps structure stories one page at a time. This demonstration showcases a test book titled "Echoes of Tomorrow" and the streamlined writing workflow.

No cloud dependencies, no distractions — just your words, your way.

Introducing CODE VAULT — a self-hosted system for securely storing technical procedures, CLI workflows, and reusable AI commands.

Narrated by ADI, this demonstration shows how CODE VAULT transforms commands like ollama list into structured, searchable entries with tags, descriptions, and step-by-step breakdowns.

Designed for engineers and system architects who need critical knowledge documented, versioned, and deployment-ready.

THE.WATCHER is a locally hosted security dashboard built to scan, classify, and monitor every device on internal networks.

This autonomous system verifies device presence before display, ensuring only confirmed systems appear in the monitoring interface. Features include real-time network visibility, Prometheus integration, and complete firewall containment.

No cloud exposure. Pure visibility and control.

VADER (Vulnerability Assessment Detection Enhanced Reasoning) is an AI-powered network vulnerability scanner designed to detect, analyze, and report potential security threats in real-time.

This educational series combines intelligent reasoning with advanced detection techniques to provide fast, efficient, and accurate assessments of system vulnerabilities.

Explore cutting-edge cybersecurity research through practical demonstrations and theoretical foundations.

SABER (Software-Assisted Broadcast Enhanced Radar) is a cutting-edge system designed to enhance radar capabilities using advanced software integration.

This system combines traditional radar technology with real-time data processing, visualization, and AI-powered analytics to provide improved situational awareness and object detection.

Ideal for defense, surveillance, and autonomous systems research with precise and reliable information processing.

In this experimental research, two local AI systems communicate using GGWave, a protocol that transmits data through sound frequencies.

Powered by GibberLink, this setup enables devices to exchange messages over ultrasonic or audible frequencies — no Wi-Fi, Bluetooth, or internet required.

This research explores alternative communication methods for AI systems in constrained or secure environments.

vendOS is a sleek, self-serve Point of Sale (POS) kiosk system designed for office or small-space vending applications.

Powered by Raspberry Pi, it combines barcode scanning, interactive touchscreen interface, and optional PayPal QR code payments — running on an honor system framework.

Users can scan items, view their cart, and check out with ease, all wrapped in a stylish, animated user interface.