👋 AMRHZ — AI Systems & Software Architecture
I'm Muhammad Amirul Hafiz (AMRHZ), an AI systems architect and software developer focused on:
- AI orchestration – Designing systems where humans and AI collaborate effectively
- Human–AI memory bridges – Creating inspectable, shared state between human intention and AI interpretation
- API-first architectures – Building systems around clean, composable interfaces
- Agent-based workflows – Developing intelligent, autonomous systems with verifiable behavior
- Cloud/serverless infrastructure – Deploying systems on modern cloud platforms
- Public, iterative development – Building in the open on GitHub with transparent progress
Current projects: AMRHZ Architecture Core, AP1 AI Orchestrator Engine, AMRHZ-AI-13 prototype, AP1-WEB-Console
Active location: Public GitHub repositories and cloud-based deployments
An early AMRHZ AI system prototype documenting:
- CSV-based memory – Simple, inspectable state storage
- Intent detection – Classifying and routing user intentions
- Self-learning module – Training improvements from interactions
- Flask API – RESTful interface for system integration
- CLI agent – Command-line interaction interface
Status: 🟢 Existing prototype
Repository: AMRHZ-AI-13
git clone https://github.com/amirulhafiz1132002-code/AMRHZ-AI-13
cd AMRHZ-AI-13
pip install -r requirements.txt
python run.py # CLI mode
python api/api_server.py # API modeA web-based workspace for AMRHZ/AP1 systems with:
- Backend – Python logic layer and API handling
- Frontend – JavaScript/HTML5/CSS3 dashboards
- Memory – Dedicated persistence and context retention
- Testing – Automated verification and test pipelines
- Dashboard interfaces – Telemetry, control, and monitoring
Status: 🟢 Existing web application
Repository: AP1-WEB-Console
A documentation and architecture repository focused on Human + AI collaboration patterns:
- Architectural layers – System structure and design decisions
- Integration points – Where humans and AI interact and connect
- Development patterns – Workflows for teams building AI systems
- Human-AI workflows – Code generation, review, optimization, decision-making
- Component organization – Modular, scalable design principles
Status: 🟢 Existing architecture documentation
Repository: amrhz-architecture-core
The AP1 ecosystem is a web-based interface to AMRHZ systems featuring:
- Dashboard – System monitoring and control
- Chat – Conversational AI interface
- Memory – Persistent context and learning
- Terminal – Command execution layer
- Agents – Autonomous task execution
- Providers – External service integrations
Status: 🟢 Available live
Access: Open AP1 Ecosystem
AMRHZ is documenting a Human–AI Memory Bridge: a shared state layer intended to:
- Preserve human intention – What the human actually wants
- Record AI interpretation – How the AI understood the request
- Track decisions – What choices were made and why
- Document assumptions – What both parties assumed vs. verified
- Enable verification – Humans inspect and validate before execution
- Capture results – Real evidence vs. claimed outcomes
Status: PROPOSED / LOCKED FOR DEVELOPMENT
Core principle:
HUMAN INTENTION
↓
AI INTERPRETATION
↓
STRUCTURED SHARED STATE
↓
SYSTEM / AI ACTION
↓
REAL EVIDENCE
↓
HUMAN VERIFICATION
↓
UPDATED REAL STATE
Key values:
- REAL STATE > UI SIMULATION – Show what actually exists
- HUMAN INTENTION > AI ASSUMPTION – Ask before assuming
- EVIDENCE > CLAIM – Verify before declaring success
- VERIFICATION > BLIND TRUST – Inspect before delegating
- FAILURE IS DATA – Learn from what breaks
📄 Full checkpoint: docs/AMRHZ-HUMAN-AI-MEMORY-BRIDGE.md
«Build → Test → Verify → Improve → Repeat.»
The approach is based on:
- 🧩 Modular repositories – Each system is independent and testable
- 🔍 Inspect before changing – Understand the current state first
- 🧪 Test individual stages – Verify each component works
- ✅ Verify results – Prove features work before shipping
- 📝 Document important changes – Track why decisions were made
- 🔐 Keep secrets out of source code – Security by design
- 🔄 Improve incrementally – Small, verifiable steps forward
The goal is not to make the system look advanced. The goal is to make the system actually work.
Across the ecosystem, the documented projects use:
- HTML, CSS, JavaScript, TypeScript
- React (in some components)
- Responsive dashboard design
- Python (primary backend)
- Flask (API framework)
- Node.js (in some services)
- AI APIs (OpenAI and others)
- Agent-based architectures
- Memory and context systems
- API-based integration patterns
- GitHub (version control and public development)
- Cloudflare Workers (serverless compute)
- Web-based environments
Technology choices vary by repository. This represents actual technologies documented across the ecosystem.
- AMRHZ-AI-13 prototype
- CSV-based memory system
- Intent detection module
- Flask API foundation
- AP1 web ecosystem interface
- Human-AI architecture documentation
- Human–AI Memory Bridge architectural design
- Multi-agent orchestration
- Enhanced memory and retrieval
- Additional interface connections
- Improved system observability
- AP1 workspace expansion
- Human–AI Memory Bridge implementation
- Broader system orchestration
- More autonomous workflows
- Additional cloud deployment paths
AMRHZ — Building the practical connection between Human + AI + Systems.
Transparent development. Verifiable progress. Real state over claims.