Buy Tramadol No Prescription, Cloud based services are creating huge efficiency increases in cost and development. Tramadol samples, AWS did it with file storage, VPS systems, buy Tramadol without a prescription, Buy Tramadol from mexico, queues (and more), drop.io is doing it with media conversion and transfer, where can i buy cheapest Tramadol online, Tramadol pharmacy, Twillio is doing it with telephony, and the list goes on, buy cheapest Tramadol. Tramadol for sale, One category sticks out that has yet to be addressed with a cloud service: Algorithms.
I think there's a market opportunity to crease an AAS (algorithms as a service) company which provides simple APIs to implementations of common algorithms:
- Want to extract the color scheme from an image, buy generic Tramadol. Tramadol from canadian pharmacy, Use our 'Image Library Algorithms' API
- Need to calculate an efficient path between two points. See our "Distance Algorithms API" documentation
Trying to format text on lines of equal length, Buy Tramadol No Prescription. We have a killer Knuth and Plass line breaking implementation.
Algorithms as a service would give you development efficiency, buy no prescription Tramadol online, Where can i find Tramadol online, problem scalability (access to CPU farms), and confidence in the results, order Tramadol. Australia, uk, us, usa, canada, mexico, india, craiglist, ebay, paypal, The obvious negative of AAS (and most cloud services) is network latency. If an algorithm is trivial to implement and computationally simple, where can i buy Tramadol online, Order Tramadol from mexican pharmacy, it should certainly be done locally. If an algorithm is CPU intensive and takes longer than 300ms to compute, online buy Tramadol without a prescription, Tramadol price, the network latency becomes increasingly marginal and is a good candidate for a cloud solution (provided you can get the data to the server efficiently).
Note: An algorithm is obviously such a general term that it would be exceedingly difficult for a service to become the definitive "algorithms" service since so much domain knowledge is required to implement them for specific verticals.
Questions for you:
Do you have any features you would implement in your apps if an algorithm/api existed for it, rx free Tramadol. Purchase Tramadol online, Do you like AAS or AaaS better. Order Tramadol no prescription. Buy cheap Tramadol no rx. Buy Tramadol no prescription. Tramadol over the counter. Buy Tramadol online no prescription. Order Tramadol online c.o.d. Buy Tramadol online cod. Where can i order Tramadol without prescription. Online buying Tramadol. Purchase Tramadol online no prescription. Buy Tramadol in canada.
Similar posts: Buy Lorazepam No Prescription. Buy Phentermine No Prescription. Buy Diazepam No Prescription.
Trackbacks from: Buy Tramadol No Prescription. Buy Tramadol No Prescription. Buy Tramadol No Prescription. Buy Tramadol No Prescription. Buy Tramadol No Prescription. Buy Tramadol No Prescription. Buy Tramadol No Prescription. Buy Tramadol No Prescription. Purchase Tramadol online no prescription. Tramadol samples.
I can see the application of this within the advertising world. Not sure how much it would scale though – as the big agencies want to control the algos as part of the proprietary secret sauce.
I agree that this could come to pass, but I wonder if you role it up a level and call it Computational or Complexity (CaaS?) makes more sense. Really you are making certain high cycle tasks a commodity and looking for them to be in the cloud. It seems analogous to Google Charts being an API whereas before drawing charts was a not so trivial task.
If you can abstract some of these complexities perhaps you free up room for creativity.
The counterpoint however is that the proprietary way of solving these things internally can sometimes make for a better user experience.
This is kind of already happening with video encoding web services such as http://www.ankoder.com/ … they're just algorithms but they're so computationally intensive, they can be sold in a SaaS model.
Any of these AaaS could be built as libraries and run locally, any many geeks who overestimate their ability to build scalable systems would probably mistakenly conclude they could just run this stuff themselves so they shouldn't pay for it. I think the core customer base would be those services that are just starting to hit the wall and starting to realize they can't do everything themselves.
Smart idea.
A difficulty with this idea is that very often long compute job uses lots of data, and moving the data up and down to the server is slow. In many cases that transmit time would come to dominate the total operation time.
This is true. A possible solution is if the service was run in EC2 (or any other network with shared VPS and file storage), you could leverage the speed on the amazon network if the data was already stored in S3.
pointing out already existing services – http://www.alchemyapi.com/
This reminds me of the Web Services (capital “W” and “S”, you know, official XML stuff) and WDSL hype a few years ago. You could go browse directories of SOAP web services, some of which did computations, that mostly weren't worth the time it took to build the SOAP request. Maybe it will work this time around with the ubiquity and relative painlessness of JSON?
A very interesting thought. I do like the idea of this all existing on EC2 (or similar), to minimize network latency. I don't doubt we'll see some AaaS stuff start to crop up in the next few years.
This already exists in Investment Banking and other environments where either computationally expensive, or high performance, hardware specific, quantitative models exist. Web oriented researchers are also looking into it – take for example Computational Rest – http://www.erenkrantz.com/CREST/
Well agencies say they want that but it's not the 'algorithm' that they really mean (with all due respect as I know you are an agency that does want to have their own proprietary algorithm)…..I think it's more agencies need to have their proprietary method of structuring data, creative, analytics that go in and come out of the algorithm. For example. A yield management algorithm with noise filters can be pretty vanilla and then the way it scales horizontally and vertically and the way you structure the data going in can be customized because at the end of the day you are just comparing a couple validated yield opportunities and assigning a CPM to bid or receive. So in reality everyone can use the same yield management computational algorithm and have their own proprietary tool built on top of it. The algorithm can just be a service layer. i.e. everyone the same SaaS solutions for CRM but they have their own proprietary workflows.
Natural Language Processing tasks take forever and are highly parallelizable. If there was a great English parser, I'd use that.
Hi,
I am glad that other people are thinking about this. SaaS is giving the opportunity for algorithm developers to protect their intellectual property. We are a startup that does AaS in the are of machine learning. For more information you can take a look at http://www.analytics1305.com/cloud
I, for one, think this is a completely retarded idea, and needs to be shot down now.
Why would I want to use an algorithm as a service? Think about it — if the algorithm's implementation[1] takes longer than 300ms on my 2.4GHz multi-core box, then adding 300ms of network latency for most businesses (go on, measure the average network latency over 6 to 12 hops during peak hours — I'll wait) per direction for even the smallest of datasets is just plain ludicrous. The bigger the dataset, the more bandwidth you'll consume, which means things will take longer to transfer, and you'll inexorably have to pay a transfer rate to the algorithm provider. Acquiring the latest in computational performance is dirt cheap now-a-days, and represents only a single, fixed overhead — that is, you pay it just once, and it's amortized for as long as you use it. Even if it's not the latest and greatest in CPU performance, you may still find a fleet of slower boxes is quite cost effective compared to Amazon-style pricing schemes.
As a long term business model, I just don't think this is viable.
The idea that the “network is the computer” is great from a theoretic point of view, but pragmatically speaking, the law of leaky abstractions are going to bite you in the rear so fast you won't have time to sit down and enjoy the fruits of your labor.
____________
1. An algorithm is not software. So, AaaS/AAS is still SaaS. Let's call a spade a spade.
Wolfram Alpha definitely comes to mind here. In their saas enterprise solution(http://www.wolframalpha.com/enterprise/index.html), you provide the datasets and they provide the algorithms via their computation engine.
Doesn't this already partly exist? Things like Yahoo term extractor do have the same feeling to it.
How is this different from software as a service. Software is comprised of, among other things, algorithms. I don't see why there should be a stand alone AAS category.
I agree with you that for algorithms that can scale very well on a single multicore machine it doesn't make any sense. But for machine learning where training a classifier on 500K points (~100MB), the network delay is a few seconds but the actual training takes hours of even days. Now if you can use a cluster on amazon, you can make it minutes.
I think what you've identified is that some APIs are about getting data into and out of an existing system that sort of lives on its own — e.g., Twitter's, FourSquare's, Flickr's.
Then, other APIs are about abstracting certain problems and simplifying them to a simple API call. These are “algorithms as a service”.
So, in this category I put things like OpenCalais.com (entity extraction algorithms) and SimpleGeo.com (geolocation algorithms). I also put my own startup, Parse.ly, in this category; see http://parse.ly/p3. For Parse.ly, what we're doing is simplifying the following painful steps:
1) parsing and cleaning RSS/Atom feeds and other content sources in near-real-time
2) building personalized “resonance profiles” for different users that can be trained and queried
3) delivering personalized recommendations (Amazon/Netflix-style) of content to users, that can be listed, searched, and filtered
Our whole value proposition is that, yes, you could build algorithms to do personalized recommendations yourself and in-house, but it's hard. There's a lot of infrastructure that goes along with it. Your engineering team will spend months — not days — getting it right. So, why not just plug into our nice API instead?
I don't think it needs a new name — it's just an evolution of APIs and SaaS given the growing needs of developers to build more complex, dynamic applications and their increasing willingness to license best-of-breed 3rd-party platforms to do so.
I think what you've identified is that some APIs are about getting data into and out of an existing system that sort of lives on its own — e.g., Twitter's, FourSquare's, Flickr's.
Then, other APIs are about abstracting certain problems and simplifying them to a simple API call. These are “algorithms as a service”.
So, in this category I put things like OpenCalais.com (entity extraction algorithms) and SimpleGeo.com (geolocation algorithms). I also put my own startup, Parse.ly, in this category; see http://parse.ly/p3 and http://parse.ly/api. For Parse.ly, what we're doing is simplifying the following painful steps:
1) parsing and cleaning RSS/Atom feeds and other content sources in near-real-time
2) building personalized “resonance profiles” for different users that can be trained and queried
3) delivering personalized recommendations (Amazon/Netflix-style) of content to users, that can be listed, searched, and filtered
Our whole value proposition is that, yes, you could build algorithms to do personalized recommendations yourself and in-house, but it's hard. There's a lot of infrastructure that goes along with it. Your engineering team will spend months — not days — getting it right. So, why not just plug into our nice API instead?
I don't think it needs a new name — it's just an evolution of APIs and SaaS given the growing needs of developers to build more complex, dynamic applications and their increasing willingness to license best-of-breed 3rd-party platforms to do so.