Juno Wong

• Led POC validation and on-site deployment of autonomous forklift systems for two major U.S. automotive OEMs, integrating AGVs into live production environments and improving material flow efficiency by 30%. • Led cross-functional deployment of Automated Guided Vehicles (AGVs) at customer sites, coordinating system integration, commissioning, and validation to ensure compliance with operational and safety requirements. • Configured and integrated fleet management systems (FMS) with customer infrastructure via industrial communication protocols (MQTT), enabling real-time AGV monitoring and dynamic task scheduling. • Conducted on-site testing and optimization of navigation, localization, and charging systems, developing environment-specific automation solutions that increased operational efficiency by 30%. • Delivered automation and AGV integration projects that achieved an average 30% improvement in operational efficiency per deployment across customer sites.

• Built a full-stack AI-powered SaaS app builder using React as frontend, Next.js as backend to generate vibe coding web apps from natural language prompts. • Integrated PostgreSQL using Neon and used Prisma ORM to design and persist user-generated project data and metadata. • Implemented secure, sandboxed app generation with E2B containers, enabling real-time preview and live URLs for AI-created applications. • Enabled user authentication, subscription billing, and credit-based usage limits using Clerk API. • Designed and deployed the system with Docker, CI/CD pipeline using GitHub Actions. • Supported dynamic app feature generation and rendering using React, Tailwind CSS v4 and shadcn/ui components.

• Developed Luckily, a cross-platform mental health social media app using Flutter for both frontend and backend, with Firebase as the real-time database and user auth system. • Designed and integrated an AI chatbot counselor using LangChain, LangPrompt, Anthropic, and OpenAI APIs, enabling empathetic, context-aware conversations based on user personality and recent emotional state. • Architected a scalable agentic AI backend enabling the chatbot to dynamically access user context, improving interaction relevance.

• Developed an EV charging management app for both iOS and Android using Dart (Flutter) for the front end, Node.js/Express for the back end, with Firebase authentication for secure user login. • Connected the app to IoT-enabled EV chargers via MQTT over Wi-Fi for real-time monitoring and control. • Used TensorFlow to build and train neural network models predicting optimal charging schedules based on inputs such as target charge levels, arrival/departure times, and charger availability. • Implemented reinforcement learning (RL) with TensorFlow Agents, using Deep Q-Network (DQN) to train an agent that dynamically adjusts charging priorities to maximize charger utilization and minimize wait times. • Integrated the trained RL agent on an Amazon EC2 instance, updating charging plans in real-time based on user inputs and charger status data received via MQTT. • Pushed optimized charging schedules back to the mobile app for user display and to IoT chargers for execution. • Achieved a 42% average increase in charger utilization during service hours.

• Built a Memory-Based Collaborative Filtering Books Recommender System using Streamlit as front-end for real-time recommendations and feedback collection. • Used the k-Nearest Neighbors (KNN) algorithm to generate personalized book suggestions based on user reading history and similar users’ preferences. • Utilized Pandas for data manipulation, SQLite for database management, and integrated data visualization tools to enhance user experience and provide insights on book trends.