AI & Software Engineer · Applied AI · LLM Systems · Full-Stack
I build AI products from model to production.
AI & software engineer owning the full path, from data preprocessing and model training to backend APIs, databases, full-stack apps, and containerized deployment. I ship systems that scale across government, healthcare, and enterprise.
01 - About
About me
I'm Patrick Saade, a French-Lebanese AI & software engineer based in Beirut, with a Master's in Artificial Intelligence Systems from EPITA Paris. My depth spans deep learning, NLP, and classical ML alongside modern LLM, agentic, and RAG architectures, all on top of a solid full-stack engineering foundation.
I take products from research to production, owning the full path from data preprocessing and model training to backend APIs, databases, full-stack web and mobile apps, and containerized deployment. I've co-founded an AI/ML department from the ground up, led AI modules in a national federal transformation program, and shipped applications now serving hundreds of users.
Whether it's a fine-tuned model, an agentic pipeline, or a React/FastAPI product, I care about software that actually reaches production and scales: privacy-aware design, solid MLOps, and clean architecture over demos that never ship.
02 / Skills
Skills & expertise
End-to-end delivery across the stack, from model training and evaluation to APIs, full-stack apps, infrastructure, and production deployment.
AI & Machine Learning
Languages
AI / ML Frameworks
Web & App Development
Data & Infrastructure
Architecture & Leadership
03 - Experience
Where I've worked
AI Engineer · Zaka / MadeBy
- Built and shipped AI-powered applications for WorkedPlace across React, FastAPI, and OpenAI, owning end-to-end delivery, now serving hundreds of external users in production.
- Developed a personalized Zoom bot on ZAKA's knowledge base that auto-joins meetings, transcribes, diarizes, and summarizes via a containerized AWS pipeline, replacing manual note-taking with minutes-fast recap emails.
- Designed technical content and Make automations for the Certified AI Consultant program (email classification, CV parsing, candidate ranking, WhatsApp chatbot), equipping ~100 consultants with deployable blueprints.
- Led a hands-on AI training in French for CIH Bank (Copilot 365 & Copilot Studio); mentored 4 interns and 4 capstone teams across concurrent initiatives.
Senior AI Consultant (Part-Time) · GlobeMed Solutions, Health Insurance
- Continue advising the AI/ML department I co-founded at GlobeMed, retained part-time as a consultant guiding delivery and technical strategy.
- Lead applied AI initiatives across Bahrain and Saudi Arabia, scope, technical strategy, and roadmaps for claims anonymization, document extraction, and translation, expanding into two new regional markets.
- Own solution design across tooling, architecture, and implementation, translating healthcare requirements into scalable AI systems that replace manual processes.
AI Engineer · Inheaden, Ministry of Human Resources & Emiratisation (MoHRE)
- AI lead on a major UAE federal transformation program, owned architecture and implementation of core AI modules, in production across the entire ministry.
- Rebuilt a non-functional forecasting process into a working self-service platform, engineering the coding agent to run on aggregated statistical data only, resolving privacy concerns by keeping sensitive records out of the pipeline.
- Architected a chat-with-database system over the national workforce database, giving analysts natural-language access via RAG, agentic workflows, and tool calling.
Machine Learning Engineer · GlobeMed Solutions, Health Insurance
- Co-founded GlobeMed's AI/ML department from the ground up, pitching a multilingual medical translation model directly to the CEO, IT Director, and senior leadership to secure funding.
- Designed a multilingual NLP system for translating Arabic medical documents using beam search, confidence scoring, semi-continuous learning, and BLEU/BERTScore evaluation.
- Improved workflow efficiency by 40%, scaled training data from 1K to 15K records, and improved model performance by 22%.
- Built an entity resolution solution mapping records between legacy and new systems, cutting debugging time by 50%.
AI Engineer · Augment Digital Facilities Services
- Built a text-to-SQL chatbot and a RAG-based assistant over user manuals, improving access to structured and unstructured internal knowledge.
- Developed the supporting vector database, LangGraph-based orchestration, API testing workflows, and technical documentation.
Machine Learning Engineer Intern · Quantmetry, Capgemini Invent
- Automated provisioning of AWS GPU EC2 environments with Terraform and Ansible for secure, reproducible data science workflows.
- Integrated Kubeflow with AWS and GCP and built an inference-serving setup with Seldon and Kubernetes for production-style A/B testing.
04 - Projects
Selected work
A mix of production AI systems, full-stack apps, and ML research from my public GitHub. Click any card for the full write-up, tech stack, and links.
Live demos
deployed, click and try themMore projects
Production work, research, and engineering on GitHub.
05 - Contact
Let's build something.
Available for freelance, contract, and full-time roles, open to relocation and remote. If you're shipping AI to production, I'd love to hear about it.