someone build

github

Jien Weng  • Peerlist

Yappy is an AI-powered LinkedIn engagement assistant designed to enhance your professional networking experience. This innovative tool leverages Google's Gemini AI and Playwright browser automation to help users stay active on LinkedIn without the need for constant manual interaction. Yappy automatically discovers LinkedIn posts and generates thoughtful, unique comments that reflect your personal style, ensuring your engagement feels authentic and meaningful.<h3>Key Features</h3><ul><li>Automatic Engagement: Yappy finds relevant LinkedIn posts and crafts AI-generated comments that sound like you.</li><li>Interactive Dashboard: The terminal-based interface allows you to monitor real-time activity, view stats, and manage settings.</li><li>Secure and Private: Yappy ensures your LinkedIn login credentials are never stored, maintaining your privacy and security.</li><li>Customizable Targets: Users can define specific topics or keywords to focus their engagement strategy.</li><li>Easy Setup: The onboarding wizard guides you through the initial setup, including obtaining a Google Gemini API key.</li></ul>Yappy is open source and distributed under the MIT License, making it a flexible and accessible tool for developers and LinkedIn users alike. Whether you're looking to boost your professional presence or streamline your social media interactions, Yappy offers a powerful solution for automated LinkedIn engagement.

Yappy: AI-powered LinkedIn engagement

Bellabeat, a high-tech manufacturer of products for women, is a successful small company with the potential to become a global player in the smart device market. Urška Sršen, co-founder and Chief Creative Officer of Bellabeat, believes that analyzing smart device fitness data could help unlock new growth opportunities for the company. This report aims to derive insights on how consumers are using their smart devices by analyzing available data from Bellabeat.

Bellabeat Data Analysis

Built a chatbot using the Deepseek R1 Distilled 1.5B model on Hugging Face to enhance AI interactions with transparent reasoning. Unlike typical LLMs, this chatbot shows its "thinking" process before generating a final response, making AI reasoning more interpretable and trustworthy. Key Features: - AI “thinking out loud” – displays reasoning steps before answering - Hosted on Hugging Face Spaces with Gradio UI - Optimized inference with FP16 quantization for efficiency - Benchmark results: Outperforms GPT-4o & Claude 3.5 in math & reasoning tasks Tech Stack: Python | Hugging Face Transformers | Gradio | Deepseek R1 Distilled 1.5B Live Demo: https://huggingface.co/spaces/jienweng/chatbot_v2 Source Code: https://huggingface.co/spaces/jienweng/chatbot_v2/tree/main

Building Deepseek R1 Chatbot

Eco Finance 🌱💰 A financial tracking platform designed to monitor and reduce carbon footprints through personal spending habits. 🌍 Overview With Malaysia set to introduce a carbon tax in 2026, Eco Finance aims to help individuals track and reduce their carbon footprint through daily spending insights. Many Malaysians are aware of sustainability issues, but visualizing personal impact is a challenge—Eco Finance solves this by linking financial transactions to carbon emissions. 🔍 How It Works By integrating with OpenFinance API, Eco Finance calculates carbon footprints using: 🛒 Merchant Category Codes (MCCs) from transactions 🌱 Emission Factor (CO2 in kg/RM) per merchant type 📊 Formula: CO2 (kg) = MCC Factor × Amount Spent Example: Spending RM100 at a store with 0.5 kg CO2/RM → 50 kg CO2 footprint. 💰 Monetizing Carbon Credits We generate carbon credits from users’ low-carbon spending habits and sell them on the Bursa Carbon Exchange (BCX). Companies like Petronas & Maybank purchase these credits to offset emissions, while users get rewarded for sustainable spending. 🌱 Encouraging Eco-Friendly Spending By rewarding low-carbon purchases, Eco Finance helps Malaysians adopt greener habits—turning sustainability into savings. 🔗 Project Prototype Developed in 24 hours at PayHack 2024, this prototype demonstrates the potential of fintech-driven climate solutions. ➡️ https://jienweng.netlify.app/blog/eco-finance/ 🚀 Impact & Future Potential Malaysia’s net-zero goals demand innovative ESG solutions. Eco Finance connects finance and sustainability, creating an ecosystem where consumers, businesses, and the environment all benefit.

Eco Finance

A revolutionary payment system using face recognition system. We've done in-person market survey with merchants and turns out people would love to implement our solution. Ultimately, we won a consolation prize in PayHack 2025, judges and audience love it.

FacePay

This dataset offered a great opportunity to perform extensive EDA and data preprocessing to improve prediction accuracy. Key steps included: •	Addressing multicollinearity with feature selection techniques to refine predictors. •	Optimizing data types and efficiently handling missing values to streamline the workflow. •	Leveraging XGBoost regression for precise sales price predictions, backed by robust data preparation. This project emphasizes the power of thorough preprocessing and tackling dataset challenges to deliver high-quality machine learning outcomes.

House Price Prediction - Kaggle Competition

I scraped and compiled a dataset containing 2,000 house price entries for Malaysia in 2025 from iProperty, aggregated via Brickz. This dataset covers key property data including township, area, state, property type, median prices, price per square foot, and transaction counts, providing valuable insights into Malaysia’s real estate market. Key Contributions: •	Developed a web scraper using Selenium to automate data extraction •	Processed and structured data using Python and Pandas •	Analyzed regional trends and provided data for predictive modeling and market analysis This project is designed for those interested in exploring the real estate market or building models based on property price trends in Malaysia.

Web Scraping - House Prices in Malaysia (2025

🏆 The Only Solo Award-Winning Project @ Universiti Malaya Hackathon 2025 ZakatDAO is a pioneering decentralized platform designed to modernize and automate the distribution of zakat, ensuring transparency, Shariah-compliance, and community-driven governance. Recognized for its impactful and innovative approach at the Universiti Malaya Hackathon 2025, ZakatDAO leverages blockchain, DAO governance, and xAI prioritization to deliver a secure, efficient, and inclusive system for zakat management.