Knowtions

Inspiration

Our inspiration stemmed from the growing complexity of financial decision-making in an ever-changing market, where businesses often struggle to interpret both their own data and external forces like news trends. We wanted to bridge this gap by creating a tool that empowers companies to make smarter, data-driven decisions, all while staying ahead of potential risks.

What it does

Knowtions analyzes your company's financial data from an uploaded CSV file and determines risks based on historical financial trends. It predicts the future performance of your stock using linear regression machine learning. Knowtions provides a statistical visualization in form of an interactive dashboard. Additionally, it delivers a weekly summary of sentiment analysis from recent news articles relevant to your company, web scraped and condensed into brief summaries, all sent directly to your email.

How we built it

We built Knowtions using Python for the machine learning component, leveraging scikit-learn for the linear regression model. The backend is powered by FastAPI and Axios, while the frontend is developed using Next.js, styled with TailwindCSS and Material UI. We used BeautifulSoup for web scraping relevant articles, Axios for making API requests, and PropelAuth for user authentication. pandas was used for handling financial data in CSV format, and Plotly was integrated for visualizing financial trends. For fast AI inference, we integrated Cerebras, and we used Palantir for advanced data visualization. The system also employs smtplib paired with a google cloud function to send weekly email reports.

Challenges we ran into

Some challenges we faced included fine-tuning the linear regression model to provide accurate stock predictions, which required significant testing. Integrating the various components—machine learning, sentiment analysis, and email functionality—was also challenging, but we managed to get everything working smoothly.

Accomplishments that we're proud of

We are proud of building a fully functional software that not only predicts stock performance but also analyzes sentiment from real-world news sources. Successfully combining financial data analysis with web scraping and delivering concise, actionable insights is something we are particularly proud of. We also managed to automate the process of sending out weekly reports, which is a feature we believe users will find incredibly valuable.

What we learned

We learned how to use a wide range of new APIs and integrate them into a cohesive system. Our team mastered the ability to work on different components simultaneously while merging them smoothly. Additionally, we learned the importance of knowing when to pivot from certain technologies to avoid falling victim to the sunken cost fallacy, making adjustments that ultimately improved the project.

What's next for Knowtions

In the future, we plan to expand Knowtions by adding more advanced machine learning models to improve the accuracy of our predictions. We also want to integrate additional data sources, such as social media trends and economic indicators, to provide more comprehensive insights. Another goal is to improve the user interface and add more customization options for the reports that users receive.

Built With

• axios
• beautiful-soup
• cerebras
• fastapi
• javascript
• materialui
• next.js
• palantir
• pandas
• plotly
• propelauth
• python
• scikit-learn
• smtplib
• tailwindcss