cho-zhang

Generating Synthetic User Behavior Data with LLM

This report presents a system for generating synthetic web browsing behavior data using Large Language Models (LLMs). The objective is to simulate realistic user interactions on a live-hosted website, enabling scalable testing, analytics, and AI model training without relying on real user data. By leveraging LLMs, the system generates diverse synthetic user profiles, behavior patterns, and interaction scripts based on the website's HTML structure and content. Each interaction is explained in natural language to maintain transparency and traceability. The project integrates AWS-hosted infrastructure with Flask, Nginx, and Selenium, enabling end-to-end deployment and execution of user simulation at scale. This approach bridges the domains of cloud engineering, frontend web development, and AI-driven synthetic data generation.

Table of Contents

1. Introduction
2. System Architecture

• [Infrastructure Setup](#infr

cho-zhang

Summary of Siri App Intents Features presented in WWDC

This note summarizes my work in the public presentation of WWDC 2024 - Bring Your App to Siri. Overall, I generated training data for the machine learning model so that the model would learn to understand what app actions are meant semantically and model could execute the right app & action to fullfill user requests.

Key Takeaways

• Make your app capable, flexible, and personalized with Siri.
• SiriKit and App Intents are the two frameworks for integrating your app with Siri.
  • If your app does not overlap with an existing SiriKit domain, App Intent is the right framework.
• Siri will take actions inside apps on your behalf.

cho-zhang

#!/bin/bash
set +e
# http://docs.openstack.org/user-guide/cli_cheat_sheet.html
#
# This script aims to prepare OpenStack Environment and create a new RHEL 7.2 instance.
# Prerequisite: must have a valid OS account;
#   must have OS RC File downloaded at current directory.

. bin/activate
sudo yum install redhat-rpm-config -y