Baraa Said العربية Message

Speaker profile

AboutBaraa Said

Also known as Baraa Hajjeh · براء الحجة

Baraa Said is a computer engineer and AI / RAG engineer who builds SaaS products end to end: he designs the retrieval pipeline, builds the product around it, and runs the server it lives on. He works as an AI Model Quality Analyst at Turing, chairs the IEEE student branch at Birzeit University, and is AWS Certified.

Send a message See his work
Portrait of Baraa Said in a dark suit and glasses
Baraa SaidAI / RAG Engineer

Computer Engineer

Builds
SaaS products, end to end
Focus
Graph RAG and Arabic AI
Now
AI Model Quality Analyst at Turing
Studies
B.Sc. Computer Engineering, Birzeit University
Leads
IEEE BZU Student Branch, as Chair
Certified
AWS Certified Cloud Practitioner

01 · StoryIn his own words

I want an answer you can trust, from data you already have, on a server I can keep running.

I'm Baraa, a computer engineer. Most of what I build starts with one question: can an AI answer from your own documents, and be right? Plain vector search often misses, so I design retrieval as a Graph RAG pipeline: the facts become a knowledge graph in Neo4j, and the model gets connected facts instead of loose paragraphs. In my experiments a rich graph brought a 7B model close to 70B quality.

A good pipeline is only useful inside a product people can use. So I build the whole thing: accounts, dashboards, billing, chat on the web and WhatsApp, admin consoles. Faheem lets a business train an Arabic assistant on its own documents. NabuX turns learning into a game of knowledge islands. Dukkan gives independent boutiques a shared marketplace with real payments and shipping.

Then I run it. Linux servers, Docker, nginx, HTTPS, backups and deploys, including the server this site is served from. Day to day I also work as an AI Model Quality Analyst at Turing, checking that AI output is clear and correct. And at Birzeit University I chair the IEEE student branch, where we bring competitions like IEEEXtreme, workshops and bootcamps to campus.

02 · ExperienceWhere he works

Paid work, research and the community he leads.

  1. AI Model Quality Analyst

    Turing

    Current · Remote
    • Validates algorithmic solutions and the clarity and correctness of AI-generated problem statements.
    • Runs automated test scripts against candidate solutions and keeps the technical metadata for each task accurate.

  2. AI Engineer

    RA Development

    Graph RAG · Remote
    • Built Graph RAG pipelines on Neo4j and LangChain with Gemini 2.5 Flash for Arabic entity extraction, compared against English pipelines on Groq Llama models.
    • Designed two-pass relationship fixing and hybrid retrieval; showed a 7B model approaching 70B quality through a rich knowledge graph.
    • Benchmarked Pinecone against ChromaDB and compared naive, graph and agentic RAG.

  3. Branch Chair

    IEEE BZU Student Branch, Birzeit University

    Current · previously Marketing Chair
    • Leads the branch's operations, technical events, competitions and community initiatives.
    • Brought IEEEXtreme 19.0 to campus: 200+ competitors coding for 24 hours straight. Represented the branch at the IEEE Region 8 Congress in Kraków.
    • As Marketing Chair, led campaigns and kept the branch's identity consistent across events.

03 · WorkSelected work

Products in production, a graduation project and research. Each one has a case study.

04 · ToolkitWhat he works with

The tools behind each step: retrieval, product and production.

AI & retrieval systems

Retrieval pipelines, vector search and chunking strategies for precise answers.

  • Naive RAG
  • Graph RAG
  • Agentic RAG
  • LangChain
  • Vector Search
  • Chunking

Databases & AI tools

Relational, vector and graph data layers, plus the AI tools that speed up research and development.

  • Pinecone
  • ChromaDB
  • Neo4j
  • SQL
  • Claude
  • ChatGPT
  • GitHub Copilot
  • Cursor
  • NotebookLM

Web platforms & analytics

Modern web experiences, performance, motion, analytics and deployment.

  • React
  • Next.js 16
  • Tailwind CSS
  • GSAP
  • Three.js
  • Lenis
  • GA4
  • Google Tag Manager
  • Microsoft Clarity
  • Vercel
  • PWA

Programming languages

Object-oriented programming, data structures, systems development and digital hardware design.

  • Java
  • C
  • C++
  • Python
  • Verilog
  • Shell Script
  • HTML & CSS

Engineering & leadership

Engineering fundamentals together with the leadership and communication that turn ideas into outcomes.

  • OOP
  • Data Structures
  • Linux CLI
  • Computer Networks
  • Machine Learning
  • Flask
  • Leadership
  • Public Speaking

05 · RecognitionCredentials and awards

06 · Q&AQuestions people ask

Short answers about who Baraa is and what he does.

Who is Baraa Said?

Baraa Said is a computer engineer and AI / RAG engineer who builds SaaS products end to end. He designs Graph RAG retrieval pipelines on Neo4j and LangChain, builds the full-stack products around them, and runs them on Linux servers he manages. He works as an AI Model Quality Analyst at Turing, studies Computer Engineering at Birzeit University, chairs the IEEE student branch there, and is an AWS Certified Cloud Practitioner.

Is Baraa Hajjeh the same person as Baraa Said?

Yes. Baraa Said also goes by Baraa Hajjeh (in Arabic, براء سعيد and براء الحجة). Both names refer to the same computer engineer and AI / RAG engineer who builds SaaS products end to end, works as an AI Model Quality Analyst at Turing, and chairs the IEEE student branch at Birzeit University.

What is Graph RAG, and why does Baraa use it?

Graph RAG is retrieval-augmented generation that stores the facts from your documents as a knowledge graph of entities and relationships, then retrieves connected facts instead of loose text chunks. Baraa uses it because it finds the right answer in messy, real-world documents, especially in Arabic. In his work, a rich knowledge graph brought a 7B model close to the answer quality of a 70B model.

What has Baraa Said built?

His main products are Faheem, an Arabic AI assistant that businesses train on their own documents; NabuX, an Arabic-first learning game; and Dukkan, a multi-vendor marketplace for independent boutiques. He also built RESCUE-Edge, a disaster-resilient edge-AI safety framework, as his graduation project, and co-authored research on detecting AI-generated voices.

Does Baraa build AI systems that work in Arabic?

Yes. Arabic is at the centre of his work. Faheem answers customers in natural Arabic on the web, WhatsApp, Messenger, Instagram and Telegram; NabuX is Arabic-first; and his Graph RAG research compares Arabic entity extraction with English pipelines.

What technologies does Baraa Said work with?

For AI and retrieval: Neo4j, LangChain, Pinecone, ChromaDB, Gemini and Groq. For products: React, Next.js, TypeScript, Python, Node.js, Prisma, PostgreSQL, Redis and Stripe. For production: Linux, Docker, nginx and AWS.

How can I contact Baraa Said?

Email Bsaid@ieee.org, message him on WhatsApp, connect on LinkedIn, or use the form on the home page. One message is enough, whether it is about an AI or full-stack role or a system you need built.

Questions?

Hiring for an AI or full-stack role, or need a system built? One message is enough.