All writing
AUTHOR
Nermien Barakat

Nermien Barakat

Systems & Applied AI Architect at TreeNodes

Most of what a system costs is decided before anyone writes code. I work on that part: how an operation actually runs, where its boundaries sit, and what a system is allowed to decide on its own.

WRITING

16 articles · Latest 22 September 2026

Prompt included

Skills, MCP, RAG and Memory: How an AI Agent Investigates a 500 Error

A post will not publish. Follow one 500 error from the server logs to a missing database update, and see how skills, MCP, RAG and memory support the investigation.

Read 6 min read
Worksheet included

Jira and Confluence: From a Business Question to a Working Feature

Follow the work. Make its logic visible. Three scenarios connect a map of current operations to shared decisions, scoped Jira work and validation.

Read
Quick check included

What is software architecture?

How a system’s parts fit together, why the design matters, and how to think about the choices behind it.

Read
Prompt included

If AI can build the software, what is the architect for?

A convincing portal demo leaves important decisions open. Follow a delivery-change request to see how architecture connects business rules, system ownership and recovery when AI helps build the software.

Read
Prompts included

How AI systems fit together

LLMs, RAG, fine-tuning, workflows, agents, MCP and WebMCP: what each contributes, when to use it, and how to combine them in a working system.

Read
Prompts included

What is Astra, and when would I use it?

Where OpenAI’s Astra fits in architecture and delivery: compare models and pricing, then work through a booking-system migration as the requirements change.

Read
Code included

The exception path is the real workflow

An invoice needs approval. A reviewer is away. A payment times out. Each needs a different next step. Here is how to design for those cases.

Read
Worksheet included

The person missing from the diagram

The diagram shows a handoff. Someone still has to chase it, check it and make it happen. How to design for the human work your system depends on.

Read
Data + prompt included

What the spreadsheet knows that the system does not

The new system is live. Why does the team still need a spreadsheet? Look at its columns, formulas and workarounds before deciding what to replace.

Read
Commands included

What happens when you press a computer’s power button?

From the first light to the login screen: how power, firmware and the operating system bring a computer to life, and where startup can stop.

Read
Code + prompt included

The C4 model: architecture at the right level

Explain a software system without putting everything on one diagram. A practical guide to C4 views, examples, AI-assisted documentation and tools.

Read
File + prompt included

Give AI agents a design reference

“Make it look modern” leaves too much to guess. Give your coding agent the colours, type, spacing and component rules it needs to build consistently.

Read
Worksheet included

Kaizen, applied to software

A small daily workaround can waste hours over time. How to spot one, try a manageable improvement and check whether it actually helped.

Read
Prompt included

Modelling the system before the code

Sales, Finance and Fulfilment may mean different things by “order”. Find those differences before they become database fields and API rules.

Read
Prompt + schema included

Giving AI a job description

Before AI reads an invoice or drafts a reply, decide what it should do when it gets stuck. Define the task, the review and the fallback.

Read
Prompt included

Software should follow the business

Follow one order, invoice or client request from start to finish. The waits, decisions and workarounds tell you what the software needs to support.

Read
SOUND LIKE YOUR WEEK?

Working through one of these problems yourself?

Discuss a project

A closer look

Swipe or scroll inside the view to explore the full visual.