Thread 1544488

Can someone please review my website:
https://alphapulse.org

Looking for macro-level criticisms about the idea as a whole. It's a slop fest rn so pointing out smaller issues is pointless at this stage.

>>1544488 (OP)
Too much padding and I'd like hyperlinks to be styled differently, though perhaps I'm just being antiquated
Tbh I can't provide much feedback since I'm not an investor.
Consider something to also show social media sentiment? I asume irrational retail investors rock the boat a fair bit.
Doesn't seem like you have any fundemental analysis tools, but maybe that's less important for options.
Have you considered the design strategy of 'put everything on one page'? Combining your macro, micro.
Is it justified to pu the news feed as the centre focus? Could you present the news in a different way? (as opposed to posts from the last day, instead present a list of active phenomenon or upcoming events. Tabulate future earnings, trump policies etc. into a schedule. Extract falsifiable hypotheses from past articles and automatically test them to give nostradamus scores to both websites/authors and posters.
Looks like a reddit/hackernews clone, how will you avoid posts being optimised for upvotes rather than quality?

>>1544552
>Have you considered the design strategy of 'put everything on one page'?
Expanding on this, perhaps investors discover useful insights better if all the information is on one page. After all they do have those crazy 6+ monitor setups.
Reminds me of broadsheets
Perhaps model your website after how good investors model their notes. Their notes will indicate what information the investors find useful. Dunno such notes are available though.

>>1544488 (OP)
I don't get the front page graphs - why aren't they accessible from any other tab? This goes for a lot of things

Also, when displaying graphs, setting the time period to 1d does nothing

>>1544599
I made the front page news feed only and moved graphs to a market tab (renamed monitor to market and am expanding functionality). I know adjusting the time period does nothing

the intent for the site is a place where it's to code, test and deploy new strategies, with the news feed being a faucet for ideas from studies and such, like pic rel. ideally you'd have this workflow that starts with a paper and ends with some coded up strategy and backtested results.

>>1544552
>Consider something to also show social media sentiment? I asume irrational retail investors rock the boat a fair bit.
this will end up on the new Markets tab

>Looks like a reddit/hackernews clone, how will you avoid posts being optimised for upvotes rather than quality?
seems like something that can be figured out later, a post-success problem.

>Doesn't seem like you have any fundemental analysis tools, but maybe that's less important for options.
they will come, but it will primarily be in the form of a dataset and code snippets / ui buttons.

I will try to digest the remainder of what you said. I modified the homepage to be just a news feed and added a markets tab where all the non-code and research stuff will go

I'd use the shit out of such a site if existed and facilitated rapid alpha research and strategy searching so im hoping others find value in it. I know it's not fully apparent what it's utility is at this stage however

also thank you for the feedback to both of you

>>1544653
OP again, so in regards to the pic I posted one of the things I want to see is users sharing notebooks, like put together a system to reproduce the papers findings, discuss interpretations with other commenters, etc.

I don't expect users to share any alpha or proprietary strategies, but communally developing a pipeline using shared tools for converting papers into live running strategies would be a great website and community imo. a public "library" of paper + interactive implementation (via Jupyter notebooks) could be developed.

the strategies should be able to be plugged up to a live data feed in the market view as well. that's essentially the vision

>>1544553
>After all they do have those crazy 6+ monitor setups.
the crowd im targeting is more programming and data/stats orientated than discretionary. with AI assisted development, it's easier than ever to get people into this and start tinkering, so im trying to lean into that as much as I can. ideally the core code is all high quality and easy for the AI to extend so they make minimal errors during development

>>1544653
I think the big thing that's missing for me is some kind of description or sentence explaining what the website exactly is. It took me a little while of clicking between tabs to understand what the point of the site even is. A get-started page would be nice.

Perhaps have a separate homepage explaining your product and what you can do it, and keep the current home as the dashboard. Every other webapp-y type of website does this kinda of thing, it makes sense

>>1544659
a bit of tangent but
>with ai assisted development
I feel like the bottleneck when it comes to data science and stats is the researcher's understanding of stats and which methodologies to apply, not their programming skill

>>1544661
I agree, but AI can help with that too, pic rel is kind of an example - letting AI see the Jupyter notebook outbook of some cell to recommend next steps in an analysis. ideally it'd be calling pre-built tools for this, so the user doesn't experience errors associated with letting the AI fully take the wheel.

>>1544660
yeah, I thought about doing a landing page but thought it'd make potential users feel like they're being sold a product too much. maybe I should have slides load if the website isn't cached or whatever that give a rundown. if you have any other ideas for how to introduce such a page or convey that information please lmk. at some point, when the page is more mature, I'll make a quick product demo webm / mp4 that runs through the features end to end and maybe stick that on the homepage..

>>1544662
man imagine the crazy stuff xoomers will do with excel+copilot

basically the site structure is
>homepage news feed
>develop tab (pic rel) is essentially an IDE and terminal interface with AI assisted development if desired
>research page is repository of data + strategy notebooks, uses code you defined in the IDE (if any) for strategy and analysis components and potentially even notebook UI
>market page is current catch-all for live market data visibility, charting, earnings dates, etc - I don't really want to compete with the sea of existing sites on this too much but the most important feature imo is being able to step through your strategy bar-by-bar or tick-by-tick for debugging and live monitoring, but I feel having a quick space for users to do some more niche stuff could help get it traction

>>1544664
what's a xoomer? a gen-x boomer? or just a zoomer?

>>1544669
yeah a gen-x'er
I don't think average boomers are doing much investing

>>1544670
yeah I feel like we're gonna see some fallout from AI accounting errors here soon enough. the next years will be interesting to watch

>>1544666
>>1544488 (OP)
for reference, this company does what im trying to do:
https://www.quantconnect.com/

but I think I can make a much nicer UI and workflow, and the page being news-first ideally makes it a stronger community

>>1544676
>>1544670
here's an example of how easy it can be, using the GUI, to run a multi-indicator grid search and look for correlations between indicators, set custom entry/exits etc. you can just button mash this on your phone.

it's sexy

FYI the newsfeed on the website is down due to hitting ngrok request limits for the month, and I don't want to self host such a site due to elevated security risk