Thanks for pointing it out. There are two parts to search.
Could you please point out, are you facing difficulty in fetching results or experiencing flickering of PDFs while loading?
I assure you, We will reduce both over a period of time. And will also work to add the chemical sector.
Hi Akash
I am accessing from UK and the network speed is pretty good when I click on news feed it is taking some time to load. I don’t know what is the underlying technology you use but one of the Popular one is APACHE SOLR , this gives google search kind of results
The delay in search option is to fetch the results once results are loaded then the navigation screen loading is little slow (not too much based on the size of the document ) . This experience also can be improved
All in all it is an excellent product.
Hi,
Thanks for replying back.
I believe, it is related to my server hosted here in New Delhi, as the News Feed get loaded instantly for me. I will research and see what I can do in this case.
Thank You 
I think you have to take CDN approach, where you have nodes at the last mile ( Zerodha Kite uses this architecture )
A very good write up on Fastly (Along with cloudflare world leader in CDN implementation )
Not related to this but the below writeup explains power of edge computing
Thank You. I am novice programmer. I will read the blog and try to implement it.
The idea you chose is brilliant, since it is content heavy , pdfs , text etc… , speed is what utmost important
The right link is here
https://softwarestackinvesting.com/fastly-edge-compute-explained/
Feature Update
1. Sentence Window -
We have created a sentence window view for each company under Document Search. This will enable the user to read the context before loading the PDFs.
Also, It can be used to track what management has said about a particular topic over time through Conference Calls, as depicted in the below image.
2. Database Size -
We have indexed the additional 2500 companies for document search, taking the total companies to around 3500. But all the PDFs are tagged as exchange filings as of now, we are developing an internal system to classify these documents.
Changed the duration filter to search between specific dates instead of months. Please note that the date for annual reports, concalls, presentations has been saved as 30th of that respective month. For example, the Date of any FY2020 annual report would be 30 Mar 2020.
3. ConCall Analyzer -
We have introduced a new section “ConCall Analyzer”, You can find details like Positive Sentences(in Green) and Negative Sentences(in Red), Guidance / Forward-Looking (in Orange), Participant list etc.
We have trained a few NLP models to classify each sentence into different categories like positive / negative / guidance. You may come across some anomalies, we are working further to train these models with bigger datasets, SOTA models, and high-end Infra.
We are working on the following things:
Kindly let me know your suggestions.
excellent work done by you sir
please share good set of KEYWORDs or PHRASES to look in ANNUAL REPORT
Hi, We are in process of compiling the list. Also, will try to reflect it in the dashboard itself.
Amazing efforts! Wishing you many laurels to come.
Just amazing the work done by you sir.so innovative…U r modest to call yourself novice
Hey…is the website not operational anymore?
Hi, We have moved it to finsight.ai.
use this link, its been update above
I have shut it down permanently.
Hi Akash, Any reason to shut it down? It is a wonderful tool you have built.
