Work

Two branches: research work and applied solutions.

GT Research Lab works in two connected ways: deep research into security, AI, privacy, and digital systems; and applied delivery for businesses that need web platforms, automation, APIs, and cybersecurity support.

Applied research and automation workspace representing selected GT Research Lab work.

How The Work Is Organised

Research creates the direction. Applied solutions turn it into practical systems.

Branch 01

Research Work

Privacy, Absolute Security, Zero Trust AI, AI firewalls, memory forensics, SOC visibility, and privacy-aware assistant systems.

Branch 02

Applied Solutions

Web platforms, AI automation, APIs, business workflows, cybersecurity defence, authorised testing, and technical consulting.

Branch 01 / Research Work

Research that investigates why modern systems fail people, then turns the findings into safer architecture.

Flagship Active Research

Absolute Security — Personal Data Protection Research

A major GT Research Lab research direction focused on protecting personal systems, mobile phones, computers, laptops, and everyday digital workflows from excessive third-party data collection, weak privacy controls, and hidden trust failures.

Active research direction

Problem

Modern devices and applications often collect more data than people understand: identifiers, behaviour patterns, location signals, device metadata, app activity, browsing context, cloud sync traces, analytics events, and advertising signals. The problem is not only hacking; it is the normalised loss of control over personal data.

What We Are Studying

We are studying how apps, operating systems, browsers, SDKs, cloud services, permissions, trackers, telemetry, and business models create data exposure across personal devices and business systems.

What We Found

Early research points to a security gap between what users think is private and what software ecosystems can observe, infer, store, share, or monetise through permissions, integrations, diagnostics, and third-party tooling.

What We Are Trying To Solve

The goal is to design practical protection patterns: privacy-aware system architecture, safer device and app configuration, local-first workflows, monitoring concepts, data boundary checks, and tools that help people reduce unnecessary exposure without breaking useful technology.

AI Security Research

Adaptive Zero Trust AI Gateway

A research architecture for safer AI model access where every prompt, output, model posture, and session context can be inspected before trust is granted.

Problem: Open-source and third-party AI systems can introduce prompt injection, unsafe outputs, misuse, data leakage, weak approval boundaries, and poor audit visibility when model calls are treated as ordinary API requests.
Found: AI safety becomes more practical when security decisions are observable, explainable, monitored, and connected to policy rather than hidden inside a black-box request chain.
Solved / Direction: The research informs secure AI workflows for businesses that need AI features with review points, audit trails, policy enforcement, and safer integration patterns.

MSc dissertation / active prototype

Security Middleware

AI Firewall / API Interceptor

A middleware concept that allows external applications to send AI traffic through a security layer before the request reaches a model provider.

Problem: Businesses want AI features, but many applications do not have a clear control point for prompt inspection, API key safety, provider routing, output review, and security logging.
Found: AI adoption becomes safer when applications can add an inspection layer without rebuilding their whole architecture.
Solved / Direction: This work supports future client systems where AI features need practical controls before being embedded into websites, dashboards, or internal tools.

Experimental research build

Digital Forensics

MemScope Memory Forensics

A forensic research prototype for ingesting and analysing memory artefacts, correlating suspicious activity, and making investigation signals easier to inspect.

Problem: Memory evidence can be difficult to interpret when process, network, DLL, command, and timeline signals are spread across raw forensic artefacts.
Found: Forensic tools become more useful when raw technical artefacts are organised into context, timelines, relationships, and reviewable signals.
Solved / Direction: The prototype strengthens GT Research Lab's security review thinking by connecting software systems with investigation and evidence workflows.

Prototype / research build

Security Operations

SOC Monitoring Dashboard

A SOC-style interface for visualising alerts, user anomalies, attack timelines, threat heatmaps, model posture events, and Zero Trust decision logs.

Problem: Security decisions are hard to trust when they disappear into backend logs without a readable operational view.
Found: Visibility is part of security; teams need to see decisions, anomalies, and context before they can respond well.
Solved / Direction: This work informs how GT Research Lab designs admin interfaces, monitoring tools, audit dashboards, and secure operational views.

Active UI module

Privacy-Aware AI

Personal AI Assistant — Aegis

A local-first assistant concept focused on useful automation, voice control, secure tool access, memory persistence, and privacy-aware AI orchestration.

Problem: Personal assistants can become powerful but risky if they ignore local control, tool boundaries, data storage, third-party exposure, and security review.
Found: Useful automation should not require blind trust; personal systems need privacy, permissions, auditability, and user control by design.
Solved / Direction: The concept connects directly to GT Research Lab's broader Absolute Security research around personal data, devices, local AI, and everyday privacy.

Concept / personal lab

Cybersecurity Education

GT CyberLearner

A cybersecurity and computer fundamentals learning platform designed to make security, internet basics, programming paths, and technical learning more approachable.

Problem: Cybersecurity education often feels fragmented: beginners jump between theory, tools, code, networking, and security terms without a clear path or interactive practice loop.
Found: Technical learning becomes easier when fundamentals are organised into paths, lessons, hints, interactive code, progress states, and repeatable practice.
Solved / Direction: GT CyberLearner supports GT Research Lab's education direction: helping people understand computer, internet, programming, and cybersecurity foundations before they approach advanced security work.

Public prototype build

Linux Learning System

Linux Odyssey

A fully local Linux learning platform with an interactive terminal simulator, RPG progression, visual learning engine, quests, labs, and mastery tracking.

Problem: Linux can be intimidating because commands, filesystems, processes, networking, permissions, and troubleshooting are usually taught separately from real practice.
Found: Learning infrastructure becomes stronger when practice, visual concepts, mastery tracking, and scenario-based labs sit inside the same system.
Solved / Direction: Linux Odyssey shows how GT Research Lab can build serious education systems and technical training tools that work locally without cloud or subscription dependency.

Public prototype build

Autonomous Systems

Autonomous Repair Agent

A cross-platform Python scaffold for observing system state, detecting issues, planning safe remediation actions, enforcing policy, and recording audit-friendly decision traces.

Problem: Automated repair can become dangerous if a system acts without understanding state, policy, verification, or a readable audit trail.
Found: Autonomous repair needs deterministic policy and traceable decisions before it can be trusted on real machines.
Solved / Direction: This work connects to GT Research Lab's Absolute Security and operations research by exploring safer ways to observe, diagnose, and repair systems without uncontrolled automation.

Public prototype scaffold

Business Automation

Auto-Man — Autonomous Affiliate Growth Engine

A FastAPI automation engine for fetching trends, scoring products, generating marketing content, tracking performance, and scaling winning content variants.

Problem: Growth and affiliate workflows can become manual, repetitive, and hard to evaluate when trend discovery, content generation, performance logging, and scaling decisions are disconnected.
Found: Business automation becomes more useful when generation, measurement, decisions, and scaling are connected inside one backend workflow.
Solved / Direction: Auto-Man demonstrates GT Research Lab's ability to design automation systems that combine APIs, analytics, AI-assisted content, and operational decision loops.

Public prototype MVP

Branch 02 / Applied Solutions

What GT Research Lab can build, improve, connect, automate, or secure for real businesses.

Applied Solution 01

Web Platforms That Turn Attention Into Action

When We Can Do This: When a business needs a professional website, portal, dashboard, booking flow, client area, or web application that customers and teams can actually use.
Business Problem: Many websites look presentable but do not create clear enquiries, structured operations, secure data handling, or a foundation for future automation.
What We Build: GT Research Lab can design and build websites, web apps, landing pages, admin panels, dashboards, portals, request systems, and secure deployment flows.
Outcome: A sharper digital presence, clearer customer journeys, better internal structure, and a platform that can grow into APIs, automation, and AI features.

Applied Solution 02

AI & Automation For Repeated Business Work

When We Can Do This: When teams spend too much time on emails, documents, leads, summaries, reporting, customer requests, or repetitive operational tasks.
Business Problem: Manual workflows slow teams down, create inconsistent responses, and make important information hard to process at scale.
What We Build: We can build AI-assisted workflows, document processors, customer enquiry handlers, lead qualification systems, reporting assistants, and approval flows with human review where needed.
Outcome: Faster operations, less repetitive work, clearer handoffs, and controlled AI features that support people instead of replacing judgement.

Applied Solution 03

APIs, Integrations & Middleware That Connect Systems

When We Can Do This: When tools, databases, websites, AI providers, CRMs, payment systems, or internal workflows need to communicate reliably.
Business Problem: Disconnected systems create duplicated work, fragile manual exports, weak security boundaries, and unreliable data movement.
What We Build: GT Research Lab can build REST APIs, backend services, middleware, webhooks, payment integrations, AI provider connectors, authentication flows, and logging.
Outcome: Connected operations, clearer data movement, reusable technical foundations, and safer integration patterns for business-critical systems.

Applied Solution 04

Cybersecurity Defence, Hardening & Authorised Testing

When We Can Do This: When a company needs to reduce risk before launch, improve an existing system, review access control, test APIs, or understand vulnerabilities.
Business Problem: Security gaps often appear in forms, authentication, admin areas, APIs, deployments, permissions, exposed data, misconfigurations, and unclear ownership.
What We Build: We can support defensive review, hardening, authorised vulnerability assessment, API security checks, OWASP-focused testing, monitoring concepts, and remediation planning.
Outcome: Clearer risk visibility, stronger technical controls, practical remediation steps, and better confidence before systems are exposed to real users.

Applied Solution 05

Research, Product Strategy & Technical Consulting

When We Can Do This: When a founder, company, or technical team needs to explore an idea, validate architecture, plan an MVP, or translate research into a working prototype.
Business Problem: Good ideas often fail because the technical path, risks, architecture, security model, and delivery scope are unclear.
What We Build: GT Research Lab can support feasibility research, proof-of-concepts, product strategy, AI security planning, cybersecurity documentation, architecture reviews, and technical reports.
Outcome: A clearer route from idea to build, with research-backed decisions, technical structure, and a realistic plan for delivery.

Applied Solution 06

Learning Platforms, Technical Training & Cyber Education

When We Can Do This: When a company, educator, community, or founder wants to teach cybersecurity, Linux, coding, AI, or technical operations through an interactive platform.
Business Problem: Technical education often fails when learners only read static material without practice, progress, feedback, or realistic challenges.
What We Build: GT Research Lab can build learning portals, curriculum paths, interactive code labs, terminal simulations, dashboards, quests, progress systems, and technical training tools.
Outcome: Better learner engagement, clearer skill development, practical technical confidence, and reusable education infrastructure.

Applied Solution 07

Autonomous Operations, Diagnostics & Repair Agents

When We Can Do This: When systems need observation, diagnostics, issue detection, safe remediation planning, audit trails, or controlled automation around operational failures.
Business Problem: Automation can damage systems if it acts without policy, verification, observability, and safe rollback thinking.
What We Build: GT Research Lab can design telemetry snapshots, policy-controlled actions, repair planning, verification workflows, dashboards, and audit-friendly automation agents.
Outcome: Safer operations, better visibility, clearer decision trails, and controlled automation that supports technical teams instead of hiding risk.

Applied Solution 08

Growth Automation, Analytics & Content Engines

When We Can Do This: When a business needs to collect trends, evaluate opportunities, generate content, track performance, and scale what works with clearer data.
Business Problem: Marketing and growth operations become messy when ideas, products, content, links, analytics, and decisions live in disconnected tools.
What We Build: GT Research Lab can build automation engines for trend discovery, product scoring, content generation, campaign tracking, affiliate workflows, analytics, and scaling decisions.
Outcome: A more systematic growth workflow with measurable decisions, reusable content operations, and clearer performance feedback.

Recent Work

Existing client systems, prototypes, and research builds in a compact view.

Private Client Work

Business Website / Digital System

A private client project focused on creating a structured digital presence and improving business operations through a custom web-based system.

Private client work.

AI Security Research

Zero Trust AI Gateway

An adaptive research architecture for safer open-source AI model access, developed around Zero Trust thinking, explainable policy decisions, and SOC-style monitoring.

MSc dissertation research and internal prototype.

AI Gateway Security

AI Firewall / API Interceptor

An experimental middleware concept that lets external applications route AI traffic through a security layer before it reaches a model provider.

Experimental research build.

Digital Forensics Prototype

MemScope Memory Forensics

A forensic research prototype focused on ingesting and analysing memory artefacts, correlating suspicious activity, and visualising investigation signals.

Prototype / research build.

Security Operations UI

SOC Monitoring Interface

A SOC-style dashboard module for visualising alerts, user anomalies, attack timelines, threat heatmaps, model posture events, and Zero Trust decision logs.

Active UI module.

Research Intelligence Prototype

GT Brain

An internal prototype focused on organising technical notes, research signals, project knowledge, and security-related documentation.

Experimental prototype.

Private Client Work

Secure Web Platform / Automation System

A private business system designed to improve internal workflows through a secure web platform and practical automation.

Private client work.

Public Prototype

GT CyberLearner

A Next.js learning platform for cybersecurity, computer basics, internet fundamentals, programming paths, and technical learning.

Public prototype project.

Public Prototype

Linux Odyssey

A fully local immersive Linux learning platform with terminal simulation, RPG progression, visual learning, quests, labs, and mastery tracking.

Public prototype project.

Public Prototype

Autonomous Repair Agent

A backend-first Python scaffold for observing system state, detecting issues, planning remediation, enforcing policy, and recording audit-friendly decision traces.

Public prototype scaffold.

Public Prototype

Auto-Man

A FastAPI automation engine for trend fetching, product scoring, AI marketing content generation, performance tracking, scaling decisions, affiliate links, and click analytics.

Public prototype MVP.