anmol

9 ms·

61.
by anmol 16y ago
This depends on your product / idea. Both Android and iPhone have a substantial install base, why not pick one and test first? If you get traction, then expand etc. Choose the platform that makes more sense for your app (capabilities, payme
62.
by anmol 16y ago
As a recently-minted PhD from a good university, there is at least some stigma associated with blogging or appealing to general population. your academic peers, determine where you get your next job, whether you get tenure, etc. You don't
63.
by anmol 16y ago
I'm a big fan of this sort of work, and also an area of ym past academic research and current startup. its evident there is a ton of consumer-facing data, which can add remarkable value to our lives.
64.
by anmol 16y ago
I like the general idea about peer-coaching, but why web entrepreneurs? IMHO such tools would be more valuable for non-web specialty skills, where RTFM isn't an option. e.g. I use you tube to learn how to do oil changes for my car. any ins
65.
by anmol 16y ago
Fair enough. We'll suck it up and do the detailed financial projections. It just sucks from the perspective of pivoting the product/tech if the first version doesn't work out as planned.
66.
Ask HN: Early angel/seed investors asking for detailed financial projections?
6 points by anmol 16y ago|5 comments
67.
by anmol 16y ago
Just to add to the other comments, don't let this bother you. Remember, your goal is to build a solid tech business, not appease YC or anyone else. Not to take away from the value YC and PG add, but there were tech companies before YC, and
68.
by anmol 16y ago
I like the UI. any advantages over google alerts?
69.
by anmol 16y ago
i.e. analyze our social life, finances, networks, contacts? Xobni, Skydeck, Mint, and many other popular examples.
70.
More companies in the 'quantified self' and 'personal datastore' space?(twitter.com)
2 points by anmol 16y ago|1 comments
71.
by anmol 16y ago
Interesting post. Perhaps a few other factors that affect this: a) sector. consumer web is very different from medical devices or energy. Can't think of any drug or medical device companies that started without a IP portfolio. a) east coast
72.
by anmol 16y ago
very briefly. We're comparing the time + effort spent crafting an application (which apparently is quite large, for first-timers) and not fitting the ask VS simply building it and approaching consumers / partners. If you have particular in
73.
by anmol 16y ago
Once we mockup our UI, would it be possible to have you give feedback / provide suggestions? Could you help us put together a user-trial with say 10-15 tech-savvy parents like you?
74.
by anmol 16y ago
pings from device forgotten at home. sorry about typo.
75.
by anmol 16y ago
With severely unbalanced classes, overall accuracy is not a good metric. So we use recall of the symptom class, and how it varies with the precision. Recall = 0.6 to 0.9. All results are based on cross-validation. There is a lot more sophis
76.
by anmol 16y ago
The challenge with 'think I've fallen' is that you constantly have to ping the accelerometer. From a battery-life perspective, that kills the smartphone. Also, its hard to distinguish those pings from For that particular application, its be
77.
by anmol 16y ago
Thanks for the kind words, and good luck to your family. We really do want to make this into a tool that can help people, if possible. My intuition is that being able to quantify and visualize 'quality-of-life' automatically, will help supp
78.
by anmol 16y ago
Trends are spotted in the first 48-hour window, in our sample. There is the big open question of validation outside the sample population of course. By precise, do you mean precision versus recall? The tradeoff depends on what you want to d
79.
by anmol 16y ago
good idea. we had an initial chat with one of the major ones. They'd like to see more traction of course.
80.
by anmol 16y ago
The 90% is a misquote, good catch. The actual accuracy is between 60-90%, depending on clusters of symptoms, and apparently that was stripped somewhere during editing. Oh well.
81.
by anmol 16y ago
The demographics that seem important (e.g. baby-boomers), use mobile phones but not necessarily FSq.
82.
by anmol 16y ago
Thoughts? We're thinking of individual-level prediction tools, but there are other ways of using this approach. e.g. at the community / city scale with mobile operator data.
83.
Ask YC: This is our tech, how can we make this into a product that people want?(newscientist.com)
16 points by anmol 16y ago|30 comments
84.
by anmol 16y ago
Hi Pg-- were you going to do this soon? we're looking to close on a small round this week...
85.
by anmol 16y ago
support the -10 for consulting + ibank + hedge fund. they're all smart guys, but watch out for the ones who have never built a product OR don't like getting hands dirty. although there are always exceptions, jeff bezos came from de shaw. i