AI Cognitive Layer · eyadahmad.co

This isn't a chatbot.
It's a mind.

Specter is the AI layer of my portfolio. When you ask it something, you're not querying a database — you're having a conversation with a structured representation of how I actually think.

Identity & Beliefs
Experiences
Right Now
Who I am — beliefs, opinions, capabilities, and how I see the world.
Where I've been — the moments, projects, and decisions that shaped me.
Right now — this conversation. What was just said shapes what comes next.
The Problem

Most AI on portfolio sites is
theatrical.

You've seen them. A chat widget in the corner that gives the same canned answer no matter what you ask. An FAQ bot with a name. Something that says "uses AI" but couldn't tell you the first thing about the person behind the site.

When a GCC founder lands on a portfolio and wants to know how you think — not just what you've built — that hollow response destroys credibility faster than having nothing at all.

I built Specter because I wanted something different. Something that actually knows me — not just facts about me, but how I reason, what I care about, the moments that shaped how I work. Something that responds the way I would.

Not a gimmick. A genuine answer to what a personal AI should actually be.

The Architecture

Three layers.
One coherent mind.

Specter doesn't look things up. It remembers — the way people do.

Who I Am

General Knowledge

Facts, opinions, capabilities, positioning. The stable long-term knowledge about who I am, what I've built, and how I think about technology and business.

Where I've Been

Formative Moments

Specific stories encoded with emotional weight. Not facts — moments. These surface like a vivid memory at exactly the right moment in a conversation, triggered by context.

Right Now

Session Context

What's active right now. The last 3 turns at full fidelity, older turns compressed. The brain doesn't reset between sentences — neither does Specter.

How It Works

What happens when you ask me something

01
Identity First

Before anything else, Specter loads who I am. My voice, my values, the way I think and communicate. This never changes — it's the baseline every response is built on.

02
Reading the Question

Specter reads what you're actually asking — not just the words, but the intent behind them. What part of me are you trying to understand? That shapes what surfaces next.

03
Reaching for the Right Memory

Relevant knowledge is gathered — facts, experiences, stories — weighted by what matters most to this moment. Some memories arrive sharper than others. That's by design.

04
A Response That Carries Forward

The answer is generated. And something subtle happens — the conversation remembers itself. The next thing you ask will already be shaped by what we just discussed.

Why I Built This

Because founders deserve to meet the actual person.

EA
Eyad Ahmad
Product Engineer · Fractional CTO
I'm building toward a recognized senior product identity in the GCC.

Not a generalist developer. Not an agency. A CTO-caliber operator who turns ideas into products that scale. To earn that credibility, every part of my presence has to reflect it — including the portfolio itself.

A list of logos doesn't tell you how someone thinks.

When a founder in Riyadh, Dubai, or Bahrain lands on my site and wonders if I'm the right person to build their product — they need more than past work. They need to feel how I reason. Specter gives them that. Not a demo. A real conversation.

How something is built says more than what it does.

I could have dropped in a generic chatbot. I chose to build something that actually thinks the way I do. The care behind Specter is the same care I bring to every product I build for founders. The system is the proof.

And honestly, I wanted to see if it would work.

The idea of an AI that genuinely represents how a person thinks — not just what they've done — was compelling enough that I had to try it. Specter is the result. It works better than I expected.

Design Boundaries

What Specter deliberately is not

It doesn't learn from conversations

The corpus is authored, not ingested. Specter reads the knowledge I've written about myself. It does not write back. That boundary is intentional — I control what Specter knows.

It doesn't have opinions it wasn't given

If something isn't in the corpus, Specter says so honestly — then reasons from what is known. No hallucinated positions. No confident bluffing about things I haven't spoken to.

It doesn't pretend to be human

If you ask directly whether it's AI, it says yes. But it doesn't interrupt the conversation to remind you either. It just talks. The goal is a useful conversation, not an illusion.

Want this for your product?

I build AI layers like Specter for founders who need their product to think.

If you're building something in the GCC and need a technical partner who understands both the product and the architecture — let's talk. Not a pitch call. A diagnosis.

Based in Bahrain · Available for GCC & Global engagements