Inspiration
The markets seem to drastically react to many of President Trump’s decisions, which in turn affect retail investors’ profits.
What it does
For the technology sector, it queries the most recent sources related to Trump and the tech field, then determines whether the stock prices of those companies will go up, down, or consolidate.
How we built it
y out web service and configure setting for it (1) Installed all necessary packages for Flask and our APIs to run. (2) Started building the backend using Python. We used the Gemini-2.5-Pro model for deeper reasoning capabilities. (3) Built the frontend using HTML and JavaScript. (4) Connected the front and back end together using JSON files and JavaScript. (5) Opened a free Render account to deploy our web service and configured the settings for it. (6) Created a new .tech domain and configured the A and CNAME records to match those of Render. (7) Waited for domain verification and SSL certification to ensure secure communication between client and server. (8) Cleaned up code, added comments for better comprehension, and submitted on Devpost.
Challenges we ran into
(1) Figuring out what APIs were, how to use them, and why their keys needed to go into .gitignore. (2) The Gemini API was not giving a parsable JSON format to feed the frontend. This was resolved by manually parsing it. (3) Figuring out how to get the Gemini response into a format the frontend could use. We converted the Gemini response to a Python dictionary, fetched only the necessary data, and then used .jsonify to pass it to the frontend. (4) The entirety of the backend was painful — figuring out how to use JSON files and format the UI was tough. We mainly used AI assistance for this part.
Accomplishments that we're proud of
(1) We created and deployed a full AI-powered web app for the first time. (2) We learned more in these 25 hours than in an entire semester of CS classes. (3) We have a publicly accessible domain that anyone can visit. (4) Our web app is applicable in the real world and, with improvements and updates, could be very useful to investors.
What we learned
(1) What APIs are, and how to use them securly. (2) How to set domain records, and how to deploy a web app. (3) How easily AI models can be implimented in code.
What's next for Wdtd
We will continue to work on this project, providing more accurate and relevant investment insights. We hope to make the UI more appealing and allow users to query different sectors outside of tech. Next, we plan to use Reddit APIs or Finnhub APIs to display current stock charts and investor sentiment related to news. Lastly, we plan to add a chat feature where users can ask investment-related questions.
Log in or sign up for Devpost to join the conversation.