As a rookie investor, i’ve always found that picking individual stocks or researching their fundamentals has been extremely overwhelming
There is so much information available across Annual Reports, Earnings Transcripts & Investor Decks- apart from all the brokerage reports & financials, you don’t quite know where to start
To solve this, I’m building Stock Alpha: an AI powered Answer Engine for Equity Research- https://stockalpha.xyz/
We help equity research analysts & investors in 3 major ways:
Asking Questions on Stocks
Generating Summaries of Company Filings
Creating Tables of Metrics & Financials
This is how the stock page for AU Small Finance Bank looks like- where you can use the above features by clicking on suggested questions or typing in your own:
If you’ve tried using ChatGPT for equity research, but disappointed with generic answers & lack of context around stock filings, you should check out what we’re building- Stock Alpha
From asking complicated questions like calculating Free Cash Flow of a stock for the last 5 years to generating an expert level investment reports, analysts at leading sell/buy-side research firms have been saving hours using Stock Alpha
Would love for the community here to try & give their feedback!
We currently only have 200+ stocks on the tool including the Nifty 50, but happy to custom add for signed up users. Thank you
Appreciate the effort in saving time in some areas of analysis, but on a personal note, I can say that, in the context of this forum, I am not too sure if members would use services like these beyond a point. Frauds have been called out by members digging into things both quantifiable and non-quantifiable, some which even fund houses couldn’t identify.
Sure- agree that the AI models we use are still getting better & prone to make mistakes. However, most of our users find value in not having to spend hours extracting information from stock filings by using this. Would love your feedback!
Great initiative. May I know which LLM and Embedding model you are using here? If you are using RAG that may be a good starting point but we finetuned foundation models would be the right next step. May I know your thoughts?
Hey- we have fine-tuned multiple AI models like GPT 4o, Gemini 1.5 pro & Claude to specifically be an expert at answering questions related to stocks
Apart from that, you won’t be able to ask questions against 100’s of stock filings like Annual Reports, Transcripts & Investor decks on other tools like Perplexity & ChatGPT
We curate these data-sets very carefully to ensure maximum breadth/depth of our knowledge base for accurate answers with source links
Hey Rohan - while I wish you the best, a cursory look at your app made me wonder whats the value add on top of what we already have via tools like Screener.
It may help if you can give a couple of concrete examples on how one can extract a fresh insight on a company versus what we get from a tool like Screener.
I mean, if I look up a certain company on Screener, on a single page view, I get almost everything that I would need including key insights, transcripts and summaries of the concalls that happen. And these summaries are quite good. All this in the free version of Screener. Ofcourse, free version means its restricted to 10-20 companies a month but the premium version in 5K/year.
Hi Sandeep- thanks for checking out Stock Alpha. While Screener is great at giving you structured tables of data & brief summaries of transcripts, our users still spent hours & weeks reading long annual reports, earnings calls & investor decks
A lot of company context & operational information, along with strategy details are captured in these documents which Screener fails to deliver
Using Stock Alpha- you have an Assistant available with you 24 x 7 that has context on hundreds of companies & their stock filings & can answer questions related to help you make better informed investment decisions
