(a) Assuming you are great developer, and understand trading very well, and know what traders normally need you are probably 3-5 years away from building something useful and re-sellable. Unless you do a gig for a trader. Which is unlikely to happen if you don't know much about trading and haven't already built a bunch of useful applications in that particular area. Also consider that most of retail traders have small accounts so they not gonna pay you 20-50k or more for a project. This is likely to exceed their net worth or trading capital. (b) If you're building for yourself to make money from trading - even longer, likely never. Best return on your time will be to you invest it into finding another SW Dev job. If you try (a) - think of your business model first and who are your clients. And learn what is out there on the market. Val
Data Mining would be great. Something that fetches global indizes and fills spreadsheets with hourly data and also adds Fed press releases to it for comparison.
hmm combining real time data with real time news ? could be interesting, i was thinking yahoo did this already (or someone) but still interesting.
I am sure,it has been done many times using various methods. And if something works it will likely not be released to public.
I thougjt more of the excel file. Im paying an Indian to do this for me, which is expensive in thw long run. Thing is I want to have staristics how specific patterns play out, how reliable price action is to cut it short. And also things like how does the market react to certain fed announcements, how long the there aformed trends last and the like. Its already paying back more then it cost but its alot of effort on my behalf (and on my wage slave's as well).
Why not a universal timeseries data converter? Processing could be client sided to avoid server costs.