ZIAD ABDALLA.Hire me

Applied AI Engineer · RAG → Knowledge Graphs · API + UI

I ship production LLM products end-to-end.

Applied AI Engineer — RAG, knowledge graphs, the API and the UI. Founder @ ZiadBrief.

87× pipeline speed-up in production
AVAILABLE · REMOTE · GMT+2 · EU / GCC / US HOURSLIVE SYSTEMS · FORGED IN PRODUCTION
02 / About

I build the whole stack of an LLM product — retrieval, knowledge graphs, the API, the UI.

Available full-time now. Bachelor thesis on RAG architecture already completed at DHBW Stuttgart; B.Sc. Computer Science & Engineering at the German University in Cairo — coursework complete, degree Jul 2027. Based in Cairo (GMT+2) with a Qatar residence permit, so remote work across EU / US and GCC hours needs no sponsorship.

LOCATIONCairo, Egypt · GMT+2
Remote EU / US · GCC hours · Qatar residence permit
EDUCATIONB.Sc. CS & Engineering
German University in Cairo · coursework complete, degree 2027
THESISRAG architecture
DHBW Stuttgart · completed 2026
LANGUAGES
Arabic · nativeEnglish · C2German · A2
Raising the archipelago…
03 / Projects
The Archipelago
Seven systems · ← → to sail · scroll on to erupt ↓
01 / 07
04 / Founder
FOUNDERZiadBrief · in development

Document AI for mid-market M&A due diligence.

Source-cited answers for data rooms. EU residency and tenant isolation were in the design from day one, not patched in later.

Visit ZiadBrief ↗

EU-only

data never leaves the EU — enforced in the architecture

GDPR-by-design

tenant isolation from the first commit

Source-cited

every answer traceable to the data room

05 / How I build with AI

I run AI like an engineering org: a frontier model plans, a fleet executes and checks its own work.

FRONTIER · PLAN
decompose · architect
MID-TIER FLEET
execute + self-verify
build fan-out
MID-TIER FLEET
execute + self-verify
(×N parallel)
MID-TIER FLEET
execute + self-verify
adversarial review council
EVAL GATE · VERIFY
golden-set · Playwright
01 · TOOLING

Claude Code power user

Custom agent skills, lifecycle hooks, MCP integrations — autonomous overnight runs that verify their own work.

02 · ORCHESTRATION

Multi-agent orchestration

Parallel build fan-outs, adversarial review councils, and self-improving playbooks.

03 · ROUTING

Model routing

Frontier models plan, a mid-tier fleet executes and self-verifies — escalating only past a model's skill.

04 · EVALS

Eval-gated development

Golden-set benchmark + AI-driven Playwright E2E. Nothing ships until the evals pass.

06 / The stack

Forged from a deep stack.

The tools behind the systems.

LangGraph
orchestration
FAISS
vector store
Claude API
reasoning
Eval-gated
eval
NebulaGraph
graph db
FastAPI
api
Redis
cache
Playwright
testing
Languages
PythonTypeScriptJavaScriptJavaSQL
LLM / RAG
LangChainLangGraphRAGKnowledge GraphsFAISSClaude API
Backend
FastAPINodeExpress
Frontend
ReactAngularNext.jsTailwind
Data
NebulaGraphNeo4jPostgreSQLMongoDBRedis
DevOps
DockerGitHub ActionsPlaywrightSentry
07 / Contact

Open to remote AI-engineering roles.

REMOTE · GMT+2 · EU / GCC / US HOURS
© 2026 Ziad Abdalla · designed and built by meBack to top ↑