discount.ai

Landing Page 1/3
Landing Page 2/3
Landing Page 3/3
Create DCF 1/2
Create DCF 2/2
Analyze/Query DCF 1/3
Analyze/Query DCF 2/3
Analyze/Query DCF 3/3
Sample DCF: Historicals are retrieved via API calls and projections are based off a rolling average.

Inspiration

We picked this idea because two of our team members have experience in the finance industry and have had the same pain point when analyzing and evaluating public companies. We have had discussions with our peers and analysts working at investment banks, and they all agree that creating models and parsing existing models is a tedious and straining process.

What it does

discount.ai automatically generates financial valuation models based on historical data. With simply a company ticker, our app will generate a custom discounted cash flow model and provide it in a downloadable Excel file, which can be further customized to user specifications. discount.ai's analysis function allows for any model to be efficiently summarized and queried by users, allowing for smoother document onboarding and communication across firm leadership.

How we built it

We built the backend using mainly Python, LangChain, and OpenAI API. This was packaged into an endpoint using Flask and called by the front-end client which was designed using React. We have two separate functions: Create & Analyze. Each function has its own endpoint that allows for a dual-feature product.

Challenges we ran into

There were several challenges we ran into such as creating endpoints for API calls, maintaining OpenAI sessions, and finding data sources. Integrating middleware was especially challenging.

Accomplishments that we're proud of

We created an MVP of a product that lets users build customizable DCFs instantly. We build an elegant front-end client that connects with our model in the backend. Moreover, our OpenAI connectivity allows for seamless training and querying of the model and external company info.

What we learned

We learned that collaboration was key to the success of our project. By delegating tasks and playing to our strengths, we were able to build something that would've taken us weeks individually. Additionally, we also learned several new frameworks such as OpenAI, LangChain, AWS, and Flask.

What's next for discount.ai

Our next steps will be to incorporate different financial valuation models such as LBOs and Comparables and allow for further user customization based on provided formatting templates, etc.