Aysun Itai

Designing and scaling creator-economy financial and tracking infrastructure powering automated payout workflows and adaptive media scraping systems. Key contributions: • Architected a multi-tier scraping lifecycle (hourly → daily → asleep) to dynamically optimize scraping frequency based on creator activity • Designed and implemented automated tier transition rules with state propagation across account-level and post-level tracked media • Built migration and backfill scripts to safely apply lifecycle logic to existing production data • Implemented role-based payout approval workflows with idempotent transaction handling and state-safe receipt generation • Improved scraping cost efficiency and system scalability by introducing activity-based tiering logic • Collaborated across product and QA to refine tracking UX and operational transparency Tech Stack: Node.js, TypeScript, PostgreSQL, TypeORM

Delivered end-to-end backend and full-stack solutions, from early prototypes to fully deployed production systems: Built an internal email campaign system using PHP, MySQL, and TypeScript for sending and tracking bulk emails (1,000+ recipients). Integrated Froala editor to allow non-technical users to create and manage rich email content. Developed a landing page and scheduling interface where users can register for workshops directly from email campaigns. Implemented tracking for delivery and user engagement (opens / interactions). Replaced the need for external tools like Mailchimp for campaign distribution.

LLMs · LangChain · MCP Tools · Chat Systems Freelance AI and backend engineer with hands-on experience building production-ready LLM systems. Specialized in LangChain-style orchestration, agent/tool architectures, and MCP (Model Context Protocol) tools for safe AI execution. Worked extensively with OpenAI, Anthropic, and Google LLMs, designing multi-step chains, routers, and chat-based AI workflows integrated into backend services. Strong focus on reliability, structured outputs, error handling, and observability in AI-driven systems. Delivered end-to-end solutions using Node.js, PHP, MySQL, and React, often integrating AI capabilities into existing products.

I teach programming to students from beginner to mid-level; JavaScript, React, CSS, Web Development, Algorithms, and Data Structures. I thought over 2000 hours and have 5-star reviews based on 70+ students' opinions.

Mentored and supported women transitioning into web development Provided hands-on guidance in HTML, CSS, JavaScript, and React

Simulation of DNA and Carbon Nanotubes in both water and vacuum; For this project I had to learn and use; LINUX FORTRAN LAMMPS-A classical molecular dynamics code with a focus on materials modelling

Built and set up an internal technical lab to support production and testing of security equipment Worked closely with production teams to support hardware validation and technical workflows Implemented Linux-based systems and tooling for testing, simulation, and analysis Gained early hands-on experience in system reliability, process discipline, and production environments