Conversational Ordering SaaS
A multi-tenant SaaS enabling restaurant ordering via an LLM chatbot on WhatsApp — with menu management, order tracking, WhatsApp Catalog and Flow.
AWS ECSDynamoDBBedrock
Available for AI projects
Fatima Ayoub
AI Engineer
I design and build end-to-end AI pipelines for NLP and generative AI — from RAG systems and LLM fine-tuning to real-time voice agents — deployed at scale on AWS.
01 — About
Turning AI research into shipped products
I'm an AI Engineer based in Beirut, Lebanon, with a background in Computer & Telecommunications Engineering from the Lebanese University. I specialize in building intelligent systems that combine large language models, retrieval-augmented generation, and real-time voice technology.
At White Stork, I design scalable AI pipelines using AWS AI/ML services and integrate OpenAI models into enterprise applications — optimizing everything from data workflows to production LLM deployments, and collaborating across teams to deliver intelligent automation and AI-powered user experiences.
2024
AI Engineer since
5+
Production AI agents
3
Languages spoken
Education
B.S. Computer & Telecommunications Engineering
Lebanese University, Beirut · 2024
General Sciences High School Diploma
Moussa & Ismail Abbas Official H.S. · 2019
Languages
- ArabicNative
- EnglishProficient
- FrenchProficient
Focused on GenAI, RAG, and real-time conversational agents.
02 — Experience
Where I've worked
2024 — Present
AI Engineer
White Stork · Beirut
- Design and develop end-to-end AI pipelines for NLP and generative AI use cases.
- Leverage AWS AI/ML services (SageMaker, Bedrock, Comprehend, Transcribe) to build scalable solutions.
- Integrate OpenAI models and APIs into enterprise applications for enhanced language understanding.
- Optimize data workflows for training, fine-tuning, and deploying LLMs in production.
- Collaborate across teams to deliver intelligent automation and AI-powered experiences.
2022
Engineering Intern
Ogero · Beirut
- Deployment of copper-based broadband services (ADSL, VDSL, HDSL).
- Deployment of fiber-based broadband services (FTTC, FTTB).
- Established the central office for network service distribution.
- Designed and managed the main FTTx network architecture.
03 — Projects
Selected work
WhatsApp Voice Calling Agent
A real-time WhatsApp calling agent built with LiveKit, Asterisk, and Pipecat for natural voice conversations.
LiveKitAsteriskPipecat
Real-Time Voice Translation
A LiveKit-based agent for real-time multilingual speech translation, supporting multiple STT, translation, and TTS providers.
LiveKitSTT/TTSReal-time
AI Trip Planning Agent
An agent-based trip planner with tool usage that generates complete itineraries in a structured JSON format.
AgentsTool UseJSON
Multimodal Fashion Search
A multimodal text-to-image search system with an agentic RAG mode that recommends outfits based on user preferences.
MultimodalAgentic RAGOpenCV
Arabic News RAG Agent
An AI agent using retrieval-augmented generation to surface relevant Arabic news articles from a large corpus via Pinecone vector search.
RAGPineconeNLPArabic
04 — Skills
Technical toolkit
AI & Machine Learning
GenAIMachine LearningDeep LearningNLPRAGOpenCV
Voice & Agents
LiveKitPipecatAsteriskWhatsApp Business API
Cloud & DevOps
AWSGCPDockerGitLinuxAgile
Languages & Frameworks
PythonFastAPIStreamlitGradio
Data & Vector DBs
PineconeWeaviateNetworking