Lamia Eid

• Developing a platform that allows developers to describe APIs in natural language and instantly generates
fully functional CRUD endpoints.
• Built a multi-layer architecture: React frontend (Vite, React Query, Zustand, Tailwind CSS) → Express
gateway (Node 20, JWT, Redis) → FastAPI AI engine (OpenAI GPT-4o, Pydantic v2, Prisma ORM).
• Implemented dynamic routing and schema registry to support user-specific APIs, with automatic Prisma
migrations and validation rules generated by AI.
• Designed and integrated PostgreSQL database with dynamic tables for users, API projects, field
definitions, and API logs.
• Enabled multi-environment deployments using Vercel (frontend) and Railway (backend & AI engine),
with CI/CD pipelines for automated updates.
• Focused on developer experience with live endpoint testing, API playground, and secure authentication
with rate limiting.
• Technologies: React.js, Vite, Tailwind CSS, Zustand, React Query, Express.js, FastAPI, Python,
PostgreSQL, Prisma, OpenAI GPT-4o, JWT, Redis, Vercel, Railway.

• Developing Glamora, a full-stack e-commerce platform with a modern, user-friendly interface, enabling
seamless product browsing, ordering, and account management.
• Built a comprehensive admin dashboard to manage orders, users, products, and categories, streamlining
business operations and improving overall efficiency.
• Implemented a secure authentication and authorization system using JWT, supporting user roles and
protected routes for both customers and administrators.
• Designed and developed RESTful APIs using Express.js and integrated databases MySQL for efficient
data handling and scalability.
• Deployed and managed the full-stack application using Railway for backend services and database
hosting, configured environment variables and infrastructure, connected a custom domain purchased via
Namecheap, and deployed the frontend on Vercel for fast, scalable delivery.
• Leveraged Railways GitHub integration and CI/CD pipelines to automate deployments, enabling
faster development cycles and reliable production updates.
• Optimized performance and responsiveness using React.js, Tailwind CSS, ensuring a smooth devices.

Platform for Sales & Marketing Teams::
• Spearheaded the development of MADs, 15
React 19, enabling hyper-personalized video outreach resulting revolutionary sales and marketing automation capabilities.
• Developed and integrated core AI-powered features including dynamic video generation based on
CRM/LinkedIn data, multi-step campaign creation workflows, real-time analytics dashboards, and
automated multi-channel delivery systems, resulting in a 50% increase in user engagement and campaign
effectiveness.
• Engineered robust frontend architecture with TypeScript, Tailwind CSS, and HeroUI components,
implementing advanced features such as internationalization (Arabic, English, Spanish, French), responsive
design, and smooth animations using Framer Motion, increasing user satisfaction by 30%.
• Built scalable component architecture with reusable UI components, context-based state management, and
optimized performance through code splitting and lazy loading, reducing page load times by 35% and
improving platform stability across web and mobile devices.

●Designed and built a production-ready, end-to-end AI-powered intake and triage pipeline for a tutoring
platform, processing unstructured user messages through a fully automated multi-step agentic workflow.
● Architected a modular pipeline: Express.js REST API (ingestion layer) → LLM classification engine
(OpenAI-compatible, Groq/Llama 3.3) → enrichment layer (entity extraction + summary generation) →
deterministic routing and escalation rule engine → structured JSON output.
●Engineered three production-grade LLM prompts (classification, entity extraction, summary) with strict
JSON schema enforcement, calibrated confidence scoring, and output validation to eliminate hallucinated
categories downstream
● Implemented a rule-based routing engine mapping ticket categories to support queues (Engineering,
Product, Billing, IT/Security) with automatic fallback to human review for low-confidence classifications.
● Built a multi-condition escalation system that auto-flags tickets based on confidence thresholds, outage
keyword detection, and high-value billing triggers, with full reason logging per escalation event.
● Designed the system for provider-agnostic LLM integration using the OpenAI SDKs baseURL
abstraction, supporting OpenAI, Groq, and Ollama with zero code changes
● Technologies: Node.js, Express.js, OpenAI SDK, Groq (Llama 3.3-70B), JavaScript (ES2022), REST API,
JSON Schema, dotenv.

• Front-End Developer Gained hands-on experience with HTML, CSS, JavaScript, and
React.js by building responsive interfaces and contributing to real-world projects. Improved code performance and
accessibility while applying UI/UX best practices.