This is an archived blog. Click here to go to the latest event.

Recently in mashups Category

Thomas Howe on Voice Mashups

| | Comments (1) | TrackBacks (0)

Click to listen

Download the audio file - 24.1 MB

Right-Click (PC users) or Control-Click (Mac users).

'; ?>

One of the most chatty, interesting and passionate people about voice that I met at the former ETel was Thomas Howe. So I figured the other week it would be good to interview him to get his perspective. Due to distance we used Skype for the interview. Tom is the CEO of the Thomas Howe Company and was formerly a CTO for a business unit at Comverse.

2008-01-11-thomas-howe-96.mp3

The run time is 34 minutes.

Below is some text to give you an idea without listening to the audio interview but it is by no means anywhere near a transcript nor full account of the interview (so please listen to it). Rather it is taken from notes I made in real-time. I've also added hyperlinks where I believe they could be beneficial to understanding.

I started by asking him if in the sphere of person-to-person communications whether voice has ceased to be the jewel - that voice is now somewhat passé' to which he responded:

yeah, yeah I think so, the message I am trying to give is that, what has happened for voice, is that we tend to think of the applications as being voice applications, so if your talking about voice applications, your taking about PBXs and conferencing, and pre-paid, PBX sort of functionalities, when, I think in reality, voice is much more powerful when it is used to enhance another application which has nothing to do with voice at all.

Tom then went on to give an example using Morisky Surveys:

if you're a patient and your going to get a prescription for drugs, there is a 4 question survey that you could take that would predict if your going to finish your course of care, finish your drugs,  and if your able to use a voice form to ask the patient as they give the drugs, if they are going to take that drugs, you can give them a much better experience of care,  and that has nothing to do with voice at all,  it has to do with making people more healthy. But by using voice you can reach these people who you could not reach before in a very controlled and inexpensive way. So I think the next wave of voice is nothing to do with voice applications but using voice to enhance other applications

Asked if voice then takes a secondary seat?

...voice could be applied to a hundred thousand applications, just none of them happen to have nothing to do with voice

Once I had a grasp where Tom was coming from a fired a more extended question at him "Do you see these applications that voice could be added to, as being new applications as in completely new or do you see this as just voice enhancing existing applications - is this new space or enhancing existing space?", to which he replied:

...I'm not taking about voice being used to solve new problems. I'm talking about voice being used to solve older problems, better.

Asked for another example Tom replied:

One good example is in the area of logistics, where you have an organisation that is responsible for managing computers , it could be vending machines, it could be computers, could be fleet management, anything to do with taking care of inventory outside of a corporate wall. One problem is tracking data and what happens to those assets. One way of extending your business process outside the firewall is by using voice to any black phone, so imagine that your doing asset repair out in the field, you could use your phone to call into the corporation , give what happened to the asset and hang up. This has some real great advantages. First of all none of the corporate data goes across the firewall. Secondly it works with every single phone. Smart phone or not. Thirdly it is very inexpensive to implement for the enterprise manager. The fourth thing is it is very easy business case to make. The fifth thing is there is really no other technology today that could give you that sort of bang for the buck. Not even web browsers because if your running them in the truck, you have the whole issue of wireless Internet access and smart phones, it's a real pain. So by using voice in the asset tracking applications, you make that problem much easier to solve, but it has nothing to do with voice.

When asked for example clients which the Thomas Howe Company works with, he replied:

I'll give you a couple of them. The first client that we are working with , actually it is a pretty neat project is for a company called Pero Systems down in Texas and with Pero we're doing a password reset for their corporate facility. So what we are doing at Pero is taking a job function  which is a guy sitting at a desk  who resets passwords when people forget them and we're replacing it with a voice script. So instead of calling a human being to get your password reset, you call this voice script. Now this not only reduces their headcount on the desk but, which is really important to them but also ensures that they have a controlled , repeatable  and consistent process of applying security to passwords...

We're also doing an application for a charity called Poverty Action, they are funded by the Bill and Linda Gates Foundation and they do research into finding ways of alleviating poverty in the third world using communications, so we did one application for them that allows, agents in the field to give micro-loan offers, in a controlled and auditable way to merchants and using any phone. It doesn't net require any smart phones, these are poor countries, but they have fraud issues and by using voice to carry that application to the field they are able to enable a much wider reach of their field workers. I can give you more examples we working with the large financials in New York City; it's a big market for us, we sort of focusing in on disease management and pharmaceuticals, and this financial stuff.

In relation to his talk he said he would bring along more examples and "hard data" from out of Forrester and other analyst groups, which say exactly how much money Enterprises save by deploying voice mashups and that the "numbers are very very impressive"

Asked if he was working with many companies APIs, he responded that he was. Asked specifically if he was working yet with Ribbit (conference sponsor) he said that he was currently learning it.

I said that Microsoft seemed to be pushing into the voice mashup direction with NetworkMashhups.com, Tom replied:

...I must admit I spend more time on ProgrammableWeb.com than the Microsoft site...Microsoft has obviously recognized the importance of mashup technology, I hope they keep an ethos and keep it open...NetworkMashups.com has 200-300 that are listed there, ProgrammableWeb has almost 3000.

Following on from this I put it to him that one of the innovation paths that BT are looking down is providing an API to their new network, known as BT 21CN, and so I asked whether or not he'd had a chance to look at the BT 21CN's API

I surely have, and I had a real great opportunity to speak with the general manager of future voice at the Sylantro global user summit last fall and he shared with me some of their work and their efforts and I'm really thrilled by what they are doing. I mean if I was to predict the future of carriers, in a world where most voice is, or at least most voice applications have mashup infrastructures, I really think British Telecom is the shining light and is doing a fantastic job. The API is very interesting, I think the fact that it is running on BTs network , you can guarantee that it will always work, it is not going to change capriciously, is a very valuable thing....it is amazing a company that size is being that entrepreneurial. 

I follow up by asking if the cost still prohibitive:

It's funny, I never think about that. I tell you why. The cost BT has for their offering is only prohibitive if you are thinking about horizontal services. If your thinking about vertical services and I go back to the Morisky Survey, the cost to the country, to the health care providers, to the people, in real money, that can be saved simply by identifying those people who will not finish their does of penicillin, makes the cost of an API infitesmal, who cares, it does not matter. Really when I am looking at these applications, the returns, the ROIs for my customers are so high and the money and the money they are saving so big and their relative volume is so small that it really doesn't matter what the costs are , given that they are somewhat reasonable

Asked if he had looked at Vodafone's Betavine and how it contrasts to the 21CN API

the great thing about the Vodafone API is that they are focusing in on location based services, which I think is a wonderful addition to what we're doing and I know there is a company here in Massachusetts called Where which is working with other carriers but really in my mind Vodafone is the leading company in enabling applications ...you know we've heard about location based services quite a bit and unfortunately for me, most of the times I hear about location based services, I hear about advertising possibilities or ...actually I am more interested in how they can help the enterprise business process and I think the fact that Vodafone has made their API so widely available, allows these mid to large companies to do their experiments to understand what sort of LBS applications make sense. One that I happen to like is workforce automation that will figures out when the Comcast technical goes into your house, how long does he stay there and when does he leave? If you can get that information, you can figure out how much to bill that customer or you could do efficiency studies, you could relate that sort of repair took that long and just by figuring out how long someone was at a certain place, is enough to understand all that data.

Asked if there is an API service out there that lets you take speech in and send out the transcript via SMS Tom said you could gang it together with a speech to email API of which there are a few and then you just pipe that into SMS.

Tom finished off by stating:

By using voice in their application they can enforce consistency that is hard to do in other ways. So they can do a consistent collection of data by using voice forms that they would not have to do otherwise.

I had to fire in just one more and asked about the discrepancy between the global reach of Internet player's APIs such as the Google Talk compared to a pretty national focused company like British Telecom, in particular questioning interactivity between someone in the States and someone on the UK. Tom replied:

depends on important localization is going to be

Get Updates

About this Archive

This page is a archive of recent entries in the mashups category.

Low Def. Video is the previous category.

miscellaneous is the next category.

Find recent content on the main index or look in the archives to find all content.