Opinion / November 2021Voicemail in Russia and abroad

“The subscriber is not a subscriber - please leave your message after the beep!” - we hear this automatic response many times and have become accustomed to hanging up, knowing for sure that no one ever checks voicemail. I, like all my friends interviewed, can’t even check it without Google! Why do operators need this strange thing? And to charge money for calls that would otherwise be free. And not only from ordinary subscribers, but also from companies that use automatic calls. Imagine a store that confirms orders not with a call center in half an hour, but with a robot in ten seconds. And some of these calls “go” to voicemail, wasting the company’s money and disrupting statistics. Below the cut is a detective story about early media, big data, machine learning and TensorFlow.

What are these “free calls”?

Telephony is already a fairly old field, with many “historical” things and technical solutions from twenty years ago. For example, monetization: operator “A” pays operator “B” for the time of a call to a phone number serviced by operator “B”. “Everyone who enters is free!” - this is from here. Operators receive money for calling their subscribers. I remember that before there were even tariffs where they paid extra for incoming calls! This solution has pros and cons. If there are approximately equal numbers of incoming and outgoing calls, then “no one owes anyone.” More incoming calls - the operator earns money. More outgoing - spends. Operators want to make money, so they do their best to maximize incoming traffic and minimize outgoing traffic. One such cost minimization mechanism is the Early Media arrangement.

Separate menu item for requests/messages

Example:

Clients call the commercial department of the pizzeria chain: Good afternoon, you called the PIZZA-M company to contact the main office in Moscow - press 1, in Kazan - press 2, in Novosibirsk - press 3, if you want to open a pizzeria in your city press 5. After pressing button 5, you are greeted by a voicemail greeting: Hello, after the tone, state your name and the city in which you want to open a pizzeria, our specialists will contact you. Clients leave an application and you can read the text of the application and see the client number in the application or in the mail.

This is just an example, you can receive any voicemail messages using a separate voice menu button.

How to setup:

Create a separate extension number that enables unconditional forwarding to voicemail. Download or read the voice greeting “Hello, after the beep, state your name and the city in which you want to open a pizzeria, our specialists will contact you” and connect a new account in the application under the extension number details.

Early Media - when the subscriber is not a subscriber

What happens when subscriber “A” calls subscriber “B” from his cell phone, who also has a cell phone?
A lot of things happen, but if we simplify it as much as possible, then operator “A” sends a request for a call to operator “B” via the SIP text protocol, and he begins to look for subscriber “B” through the towers (in fact, via SS7 over PRI, but let’s not talk about sad). So that subscriber “A” does not have silence on the phone at this time and can sell all sorts of “replace the dial tone”, the operators agreed on the state of “Early Media”: while operator “B” is looking for his subscriber, he can answer “early media” via SIP " and start transmitting sound via RTP. Beeps, music or “sorry, the subscriber is not a subscriber.” The operators also agreed that “early media” will not be charged as an incoming call; operator “A” does not pay operator “B” for this music or beeps. And so that no one cheats, we also agreed in the “early media” state to send the sound only to the caller and terminate such a call after 60 seconds. Although even with such restrictions there are craftsmen who do something useful in early media on “free” 8-800-, but this is a different story. And our story is about voice mail.

The passing era of voice communication


One of the main capabilities of smartphones is still voice communication; after all, this device is first and foremost a telephone, and only then a pocket computer.
Although it cannot be denied that the telephone function is increasingly beginning to resemble an atavism. Perhaps this statement seems strange to you, how can a mobile phone be an atavism? There are several reasons for this assessment, and the main one is the psychological discomfort that most of us experience to one degree or another during a telephone conversation. Don't be so quick to disagree - we'd like to draw your attention to an interesting article published on Gizmodo. Yes, once everything was different. Voice telephony served a specific purpose: providing instant communication. And this completely changed the rules of the game all over the world, starting in 1844. More than a century and a half has passed since then, and during this period, enormous by the standards of technological progress, only the technical basis of telephone communications has changed. But the principle itself remained virtually unchanged. And there were all the prerequisites for this, given the fact that the very opportunity to communicate by voice with a person on the other side of the country or abroad was as accessible as it was important and useful. Mobile phones have made a special contribution to improving our quality of life. One of the main reasons people buy cell phones today is for emergency communications. But now there are many other developed methods of mobile communication: SMS, email, instant messengers. Today, the phrase “mobile phone” means a smartphone, and not its push-button predecessor. In fact, an entire “mobile culture” has already formed, for which it is uncharacteristic to use such a seemingly simple and harmless voice communication - please note that in the vast majority of cases we text, rather than call other people.

There are reasons for this. It is easier and easier to type a text; it is a less formal way of communication, in which you feel more relaxed. The call requires a certain emotional tension and composure from us. Sometimes there is pressure on the interlocutor. We live in a world where we have to plan our future calls. We usually call when there is some urgent matter that is worth bothering with a call. And often this is some kind of unpleasant urgent matter. “I wasn’t expecting a call, did something happen?”

In fact, now any other method of sending a message is more preferable than calling. In addition to the mentioned SMS, mail and instant messengers, there are also all kinds of social networks. Never before have we had access to such a variety of communication channels, with their unobtrusiveness and wide choice of visual expression of all kinds of emotions. And this is wonderful, because among all this celebration of life, a call is perhaps the worst communication option. When you dial someone's number, you don't know whether the subscriber is busy, and he doesn't know why you're calling, so the very fact of the call makes him wary and tense. Every time we call someone, we open Schrödinger's box, wondering if it's appropriate to have a conversation right now. In addition, calls are much more emotionally charged, which is why we are so irritated by calls at inopportune times or interrupting our work.

As soon as you answer someone’s call, the conversation immediately loses all those nice social nuances characteristic of personal correspondence (sly winks, shy smiles, etc.), and turns out to be filled with everyday and stupid shuffles: “Hello, how are you? Sorry, bad connection. What are you saying? I don’t understand, repeat it again. Oh, no, it wasn’t me who laughed, it’s one of the 193,940 people who are now around me in this place. Sorry, didn't you hear? Yes, bye." The worst part of a conversation, bringing nothing useful or good.

That's why more and more people are using smartphones for text communication rather than voice communication. The Swype typing mode on the on-screen keyboard, coupled with automatic word selection, makes non-voice methods of communication even more convenient and informative. Some of them are good for communicating with society, some are good for private conversations, but all of them are psychologically more comfortable and tactful than a phone call.

If you need to hear someone's voice - your family, friends or loved ones - then you usually find a secluded place to make a call. And in these cases, it is better to communicate via video chat in order not only to hear, but also to see the person important to you.

If you find yourself somewhere in a crowded place, and you are overwhelmed by the desire to share something fleeting, meaningless thoughts, just take out your smartphone, send a couple of lines - and that’s it. Modern chat interfaces allow us to constantly exist within the flow of information from our social connections.

Not having to internally mobilize when receiving an unexpected call makes communication more comfortable. That’s why today we often preface our calls with a text message, “Can I call you now? Or when is the best time to dial? This has become not so much a gesture of politeness, but a necessity. As one person aptly noted:

“Every phone call has a victim - the person who is called.
You are constantly distracted from some task or activity when they call to chat. Particularly unbearable in this regard are drivers stuck in traffic jams and calling their friends and acquaintances out of boredom. ...When I receive a text alert about a call, I am always happy about it, even before I read the message. And when my phone rings unexpectedly, I think: “Holy shit, who else is there?”

***

According to research agency Nielsen, the average length of phone calls reached an all-time high in 2007, then quickly fell and stabilized.

We have reduced the time spent with the tube in the ear to the minimum possible, and now all that remains is to abandon it altogether. True, at one time the same thing was prophesied for books, newspapers and e-mail, happily predicting all sorts of periods of dying for them. But they still somehow manage to survive. Phone calls also show all the signs of fading into oblivion. Maybe it’s time to show mercy and everyone should stop calling unless absolutely necessary and answering calls?

Of course, the telephone function in a smartphone should be preserved just for cases of various emergency and urgent situations. You won’t, while running away from a gang of painted killer clowns in the forest at night, stop to send a panicked text message to your friends. But, fortunately, there are not many situations in our lives that require immediate communication with other people. So isn’t it better, to protect each other from unnecessary worries and inconveniences, to switch to text communication?

Have you ever noticed a little tension in yourself when you call someone or when they call you unexpectedly? Do you often accept an SMS or a message in a messenger instead of a call if the matter is not urgent?

Voicemail as an "honest" way to take money

If the operator did not find his subscriber, then he did not earn money on the incoming call. Telecom operators, like any commercial organizations, love to earn money, so the ingenious “voice mail” was invented. The phrase “leave a message after the tone” gives the receiving operator the opportunity to “accept” the call even when the subscriber is not available. Honestly, record 20 seconds of silence somewhere and, most importantly, take money from the calling operator for it. The most cunning ones don’t even wait for the “beep” and immediately answer the call - why waste money?

How to enable options

Connection methods are presented in the table:

Voicemail typeCommand to send a requestConnection via SMS notification to number 111
Basic*111*2919#Send text 2919
Standard*111*90#Text "90" space and number "1"
Plus*111*900#Write the text 90, "space" and "9"

The activation procedure can be performed in your personal account.

What a person can’t do is a disaster for a robot

As a rule, cellular subscribers do not receive voicemail.
For me personally, it makes no difference whether the handset says “the subscriber is temporarily unavailable” or “the subscriber is temporarily unavailable, leave your message after the signal.” I, like everyone I know, will hang up on the word “unavailable.” And what pennies one operator will pay another for such a call is not very interesting to me. It’s a completely different matter if I’m Voximplant and our platform is used to automatically confirm an order in an online store. Early media is also free, but for voicemail, money will be taken from the client’s account at the rates of the operator whose phone the call was made to. The amount in itself is small, but multiply by thousands or tens of thousands of calls per day - and it’s not so small anymore.

But automation is not limited to “calling after the buyer has clicked the “buy” button on the retailer’s web page and offering to press one or say “confirm” to confirm the order.” There are automatic notifications about, for example, a ticket to a concert. Statistics show that the subscriber received a call and listened to the message - but in fact, the voicemail “listened” to the message. Or even worse: the automation calls clients to, for example, discuss the terms of the ordered house cleaning. She synthesizes to the client “hello, this is a robot from such and such a company, I’m calling about a cleaning order, I’m connecting with the operator”, to the operator she synthesizes “we got through to such and such a client” and shows the order card in the CRM, and then the operator talks for 20 seconds in silence voicemail.

What is VoLTE

A way out of this situation promises to be a new product on the domestic communications market – Voice over LTE (“voice over LTE”), or VoLTE for short. The introduction of this communication format eliminates the dilemma: to make a call, there is no need to switch to 2G/3G - everything happens directly on the LTE frequency. That is, voice communication is carried out via high-speed LTE Internet (4G), similar to how it is done in Skype or Viber. Some will understand the analogy with the SIP IP telephony protocol.

This allows you to remain on the 4G network even during a call, since voice traffic is now transmitted digitally along with other received and transmitted data (photos, videos, web pages).


First attempts to identify voicemail

We have been automating telephone and video calls for a long time, so we began solving the problem of identifying voicemail several years ago.
What do all voicemails have in common? They all have a "beep" that's between "leave your message after the tone" and transferring the call from "early media" to "accepted." The bad news is that “p-i-i-i-i” is different for everyone. One beep, several, on one frequency, two, of different durations and frequencies. Moreover, operators like to change this “pi-i-i-i” from time to time. I wonder why?.. Our first implementation used the Goertzel Algorithm to calculate the “carrier” frequency and heuristics to recognize the voicemail audio signal by the appearance of the frequency in the audio stream. Unfortunately, this method, although it worked, had serious drawbacks. If the operator changed the sound signal pattern, then the heuristic “broke” and we had to manually update it with the new “pew-pew-peep-peep”. Much worse were false alarms: “tricky” signals on two frequencies at once were difficult to distinguish from a human voice and showed voice mail where a live person actually answered. Customers wanted reliability.

How to listen to a voice message on your phone

You can listen to it in the ways described above. If there is mail, then the number will automatically turn it on, and the name of the subscriber who sent it will also be indicated. Through your personal account, you can click on the “On” icon to activate listening.

Also, through the newsletter, you can click on the link that is attached to the SMS and listen to the information using an Internet connection.

Deep Learning. Everywhere Deep Learning

Having failed with regular math, we decided that we should try matrix multiplication. After all, this is not just mathematics, but Deep Learning and Artificial Intelligence! TensorFlow was installed and work began: recordings of conversations and voicemails were fed to different models in the hope that they would find patterns invisible to us: characteristic time delays, even intonation, a certain set of words, all this. The first problem happened with the data: even a few seconds of voice with a “telephone” frequency of 8 kilohertz is tens of thousands of values. And the more complex the data on which we train the neural network, the more data is needed for an adequate result. To train a neural network on raw data, we would need labeled recordings of millions of
calls.

Therefore, the data needed to be processed. We connected specific telecom libraries to Python, written in C/C++ and implementing the logic of working with voice: noise reduction, echo cancellation, carrier selection and many others. After processing, the recording turned into a set of parameters on which the neural network was already trained.

The result immediately became much more fun, and for the next six months we played IT alchemists: we selected a model, options for processing input data and the results of applying the model, so that as a result, we could determine voicemail from a few seconds of recording. The result was very good - now it’s enough to unemotionally start a conversation with the phrase “The subscriber is temporarily unavailable” to receive a notification that there is most likely a voicemail on the other side of the receiver. And each client decides what to do next with the information received in cloud JavaScript. For a programmer, using the detector looks like this:

MTS answering machine operation options

After we figured out what “Welcome to the MTS answering machine” means when you call. We also list the three main forms of voice mail implementation from the mobile operator MTS. They are as follows:

  • Voicemail (basic) is a popular free form. Messages left by callers are stored for one day; the duration of the message left cannot exceed 1 minute. You can leave no more than 15 voice messages during the day;

The service is activated using the command *111*2919#; or via SMS 2919 to number 111. You can also activate the service through your personal account on the MTS website (section “Service Management”) or the corresponding mobile application (Services - Paid).

Disabling the service is done using the command *111*2919*0#, or SMS 29190 to 111, as well as using your personal account and mobile application.

  • Voice mail is a more advanced paid analogue, for the use of which 2.3 rubles are charged daily. Includes the ability to store received voice messages for 7 days (not listened to) and 10 days (listened to), the maximum recording duration is 1.5 minutes. And the maximum number of messages left for a subscriber per day can be 20 entries. A nice bonus here is the replacement of the template greeting with your personal greeting, which you can record yourself by calling 086. You also get access to your mail using the web, MMS, and email;

The service is activated using the command *111*90#, or SMS 901 to 111, through your personal account or using a mobile application (Services - Paid). At the same time, your mailbox will receive an access code that allows you to maintain the confidentiality of your records. You can listen to received messages by calling the above-mentioned phone number 0860.

Description of the “Voice Messages” option

Often the caller is unable to answer the call while at work or while driving.
Also, the smartphone may run out of power or lose the network. In order not to miss important calls, operators offer their subscribers a voice message function, which works on the same principle as the old answering machine. After basic setup, all calls that are interrupted or not accepted by the subscriber for reasons beyond his control will be redirected to a special service that greets the caller and offers to leave a short voice message. You can then listen to it and become familiar with the purpose of the call before making a call back.

Types of devices

VoIP gateways can have a different number of ports, line capacity and different designs (desktop, wall-mounted, rack-mounted, with built-in or external power supply). Based on the type of connectors, devices can be divided into three groups.

Analog gateways, through the FXS port, work as adapters to which telephone sets are connected and convert the digital signal to analog. Another port, FXO, connects regular telephone lines from the city telephone network, which provides switching of calls from the public telephone network to telephones on the IP network. The capacity of analog VoIP gateways is from 2 to 24 lines. If there is a shortage of ports, hybrid models are used, in which the analog line is connected to FXO, and the PBX port is connected to FXS. These devices are widely used due to their versatility, compactness and ease of operation.

Digital gateways are designed to connect one or more digital lines ISDN, E1, T1 to a virtual PBX or to the gap between a city telephone network and an office PBX with a capacity of 50 lines or more.

GSM SIP gateways are devices with SIM card slots that allow you to connect mobile phones to a PBX and an IP network.

Rating
( 1 rating, average 4 out of 5 )
Did you like the article? Share with friends:
For any suggestions regarding the site: [email protected]
Для любых предложений по сайту: [email protected]