Los Angeles, California, United States
Architected and deployed a production-grade Swarm Agent framework that dynamically spawns, orchestrates, and retires hierarchical AI agents to autonomously synthesize complex life sciences research.
Engineered a model-agnostic Chain-of-Thought (CoT) reasoning agent with dynamic tool-calling capabilities, successfully integrating over 100+ specialized enterprise tools (PubMed, FDA databases, ClinicalTrials.gov).
Built scalable Retrieval-Augmented Generation (RAG) pipelines from scratch, utilizing OpenSearch and KNN vector search to index and query highly technical, domain-specific knowledge bases.
Designed a secure, centralized integration layer using KMS to dynamically provision, manage, and execute credentials across Google, Notion, Oracle (OCI), and Slack.
Owned the end-to-end cloud infrastructure, deploying an intelligence service and vector databases from zero-to-one on Google Cloud Platform (GCP).
Los Angeles, California, United States
Architected and shipped an AI productivity platform adopted by 4,000+ VA employees.
Designed a Retrieval-Augmented Generation (RAG) pipeline indexing 12 M+ documents. Blended FAISS IVF-PQ embeddings with BM25 hybrid search; tuned latency to < 1.2s p95 while sustaining 92% recall.
Built cloud infrastructure as code (Terraform + GitHub Actions) covering Azure Blob Storage, Cosmos DB, Functions, API Management, and Azure OpenAI
Led full-stack feature delivery in Node.js, React, and Python; introduced pre-commit hooks & automated security scans that reduced PR cycle time by 30%.
Mentored a 5-engineer pod on best practices for vector search, prompt engineering, and IaC, and collaborated daily with product, infosec, and VA stakeholders.
Los Angeles, California, United States
Directed the development of a high-performance Middle Office Clearing platform, automating thousands of transactions daily for clients managing billion-dollar portfolios.
Built a fully automated Backend-for-Frontend (BFF) with Node.js and GraphQL, reducing frontend staffing needs by 50%.
Integrated a permissioning system with Active Directory and Keycloak in collaboration with DevOps and Architecture teams.
Automated ACAT reviews and transaction reconciliations, enhancing efficiency for Clearing staff.
Los Angeles, California, United States
Employee #8, helped scale the engineering team from 5 to over 200.
Architected and developed the company's frontend platform with React, TypeScript, Node.js, and GraphQL.
Created a custom GraphQL adapter, cutting API development time and standardizing microservice gateways.
Built a scalable PDF generation service with AWS Lambda and pdfkit, optimizing cost and performance.
Migrated codebases to GitLab and streamlined CI/CD processes, increasing release frequency.
Developed a design system with Storybook and Figma, enhancing UI consistency and efficiency.
Mentored 30+ engineers, established coding standards, and implemented SOC-2 compliance.
2018 — 2019
Greater Los Angeles Area
Built a full-stack solar farm design and cost estimation platform with Next.js, React, Apollo GraphQL, Node.js, MATLAB, and AWS Lambda.
Transitioned server-side code execution to a serverless setup, cutting AWS costs by ~80%.
Maintained legacy SaaS products with Angular.js, Node.js, and MongoDB, ensuring smooth operation.
Education
2013 — 2014
American University of Beirut
Medical School Coursework
2013 — 2014
2010 — 2013
American University of Beirut
Bachelor's Degree
2010 — 2013