PintarSaham.AI

• 

  Select Ticker

• 

  Interactive Chart

Inspiration

The inspiration behind creating an AI for an Indonesian stock prediction application lies in the goal of empowering investors with advanced tools and insights to make informed decisions and achieve potential financial growth through investments. The increasing demand for accurate predictions in the Indonesian stock market, coupled with advancements in AI technology, presents an opportunity to offer data-driven analysis and precise forecasts. Focusing on the significant LQ45 index, which includes influential companies, attracts investors interested in large-cap stocks. Moreover, by providing affordable and user-friendly access to stock market insights, the application can bridge the gap for a wider range of investors. Additionally, the application serves an educational purpose, allowing users to understand AI algorithms and their applications in financial data analysis. The potential social impact of aiding investors in making better financial decisions adds to the motivation for the project. However, users are advised to use AI predictions with caution, understanding that the stock market remains unpredictable, and supplementary analysis is crucial.

What it does

PintarSaham.AI is an innovative and cutting-edge project that revolutionizes stock market analysis by leveraging the power of Large Language Models (LLMs) from OpenAI. This groundbreaking platform aims to provide investors and traders with comprehensive insights into the current market sentiment about a specific stock while offering accurate forecasts of its future price movements and monthly returns. Through the seamless integration of three Long Short-Term Memory (LSTM) models for optimistic, pessimistic, and average predictions, along with the implementation of the renowned Fama-French 5 Factor model, PintarSaham.AI opens up new frontiers in informed decision-making in the financial world.

How we built it

We design user-friendly app using the stocks website as our reference. Built it using Express for the server side, Next.js for the client side, and Tensorflow integrated with Flask API for the prediction model. Find insightful news and forecast according to various stocks in Indonesia and utilize it for hassle-free investing!

Challenges we ran into

• Implementing and integrating database to the backend
• Setting up MVC pattern from plain Express framework
• Converting Math Foundation of LSTM to Python code

Accomplishments that we're proud of

• Learnt how to work efficiently in a streamlined time
• Collaborative Coding for the first time
• First Experience Hackathon
• First Machine Learning App Development

What we learned

• Machine Learning Deployment
• Multi-framework Development
• Interactive Chart and Websites

What's next for PintarSaham.AI

We intend to widen the stocks list from LQ45 to all IHSG stocks. Furthermore, we aim to create Reinforcement Learning Model for guiding the investors when to buy or sell the stocks according to the rewards & punishment based on the action that the machine takes.

Built With

• express.js
• next.js
• prisma
• python
• tensorflow