Any plans for other languages and locales? I immediately noticed the temperature in F in the example about the weather in Lima. I think everybody there uses C with the exception of American tourists :-) Seriously, it looks a great product. Maybe it returns even too much data in the JSON. I wonder how to take advantage of all of that if I don't know what people are going to ask. They're going to ask silly questions just for fun even if I have a vertical app (example: a mortgage calculator), because this is not a web form with constrained input fields but a free form input. The numbers I get into the answer could be unrelated to mortgages. Do you have examples of best practices? Maybe just write and speak the answer? Thanks.
You're probably skeptical (as I was) but watch this video demo of the Hound app: <a href="https://m.youtube.com/watch?v=M1ONXea0mXg" rel="nofollow">https://m.youtube.com/watch?v=M1ONXea0mXg</a><p>That's insanely fast, compound natural language queries. I'm impressed.
After owning Echo, Roku and Fire TV, I'm super-bullish on voice commands finally being ready for prime time. It's a terrific interface for home audio, TV and car audio.<p>I've gotta think Apple will open up Siri to app developers sooner than later.<p>Houndify looks interesting.
It's already really easy to get fast, efficient access to large data sets. I don't see much value in that. It is not fast,efficient, and easy to transform natural language queries into computationally actionable ones.<p>I would find more value as a developer if, when given a natural language query, it returned a structured query. Then I could tweak the query to conform to whatever data retrieval API I wanted.<p>I don't think what I'm asking for has to be mutually exclusive with what they're currently offering. Give me the option to have houndify do some or all of the work for me.
Also <a href="https://news.ycombinator.com/item?id=9650748" rel="nofollow">https://news.ycombinator.com/item?id=9650748</a>.
I've been using pocketsphinx with this neat Ruby gem[1]. It's really easy to use but has low accuracy (understands me correctly maybe half the time). I'm curious to see if Houndify does any better!<p>[1] <a href="https://github.com/watsonbox/pocketsphinx-ruby" rel="nofollow">https://github.com/watsonbox/pocketsphinx-ruby</a>
There is clearly a knowledge graph coupled with this in addition to the speech recognition. Sorry, "meaning" recognition. I feel like there is an opportunity to connect the deep knowledge graph of Wolfram Alpha -- or that maybe Wolfram missed the ball by not connecting their graph in a more usable way.
Does this require a network connection? I'd love to start adding speech-to-text interfaces to my apps, but most of the stuff I work on needs to be able to work without the network, and most of the speech-to-text engines these days are SaaS products in some form or another.
Can't Android Google voice keyboard be used as a speech-to-text interface and then the text can be used to trigger a command in a similar fashion?
This looks extremely cool and I can't wait to try it.<p>Bug: When scrolling down the page it is very very sluggish, using Chrome on Xubuntu 15.04.