|  telisma news
|
teliSpeech news |
telisma's customers & partners |
|
 telisma raises an additional €1.2 million with UFG Private Equity to strengthen its international operations
|
|  | Paris, May 29th 2007 - telisma, European leader of speech recognition technologies announces its increase in capital of €1.2 million. The new placement was led by UFG Private Equity and followed by its historical shareholders: 3i, Ventech, Innovacom and AGF Private Equity. This capital increase comes at a time where the market is showing a strong demand for speech understanding and interaction solutions across the globe. It will allow telisma to take full advantage of this momentum with its teliSpeech platform.
Laurent Balaine, CEO of telisma says: ‘With the teliSpeech platform we make it simpler, faster and cheaper for partners to make high quality speech solutions in different languages. With the funds that we’ve raised, we will accelerate our development into international markets such as in India and continue to strengthen our product leadership’.
According to Olivier Denigot and Guillaume Toussaint, Investment Directors at UFG Private Equity: “Thanks to its development tools and its experience of mass market applications we believe telisma is now ideally positioned to take significant market share in international markets. In addition, the growing use of unstructured data makes speech recognition a must and should become an important growth driver for telisma in the near future”. |
high performing speech recognition in Indian English and Hindi available now with teliSpeech |
| | teliSpeech Indian English and Hindi unique features!
| - first ever Hindi grammars written in Devanagari script!
- all Indian English accents taken into account for unmatched accuracy
- Windows and Linux available
- more Indian languages to come soon
|
Developed from scratch and using native speakers, telisma’s Indian languages include their own phonetics and related acoustic libraries and they support the appropriate Indian script.
The Hindi language provides local application developers the unique ability to deal with dynamic grammars for contents written in Devanagari script, telisma is proud to be the only ASR engine to offer such a key feature. The Roman alphabet is also available.
The Hindi acoustic library has been trained on large speech corpora of Hindi speakers from various regions in the country.
The Indian English deals with the great variety in Indian people’s accents when speaking English. The English accent being understandably influenced by the speaker’s mother tongue, so telisma uses speech databases created by native speakers and collected from all the States in India. India as part of telisma’s international development strategy |
With a booming telephony sector, India is seeing an increasing demand for speech access to content. With no less than 18 languages officially spoken nationwide – not to mention the tremendous variety of dialects commonly used in rural areas – India is undoubtedly one of the most challenging countries for Speech Recognition! True to its international focus, telisma will be releasing Indo-Aryan and Dravidian languages later this year.
For further information please contact Richard Mellor on +44 1403 701889 or come meet us at the VAS show in New Delhi on the 6th July 2007! |
Orange Business Services deploys at Generali an automated attendant based on natural dialogue relying on telisma’s technology |
| | Generali Gestion Santé, the entity in charge on managing health expenses for Groupe Generali’s French customers, selected Orange Business Services to design a new customer relation management telephony system.
Generali chose an innovative system based on natural dialogue; an automated attendant welcomes callers, and easy questions, such as hours of operation, pricing, etc., are answered automatically. The caller can express himself naturally: the automate interprets his demand - whatever the words or expressions he uses – and is able to dialog shortly with him.
Generali is the first insurance company to install such a system. Henri Laurent, Generali Gestion Santé Director says: “The quality of customer’s service is a priority in the health insurance sector, where questions arising are numerous and manifolds. The telephone is the main entry point for our subscribers, and its management is a key performance factor for our group. Thanks to Orange Business Services advices, the number of calls successfully handled and call qualification have been significantly optimised” | |
| about Generali |
Generali Gestion Santé belongs to the Generali group created as early as 1831 in Italy.
Generali:
- is the third largest European insurer,
- is present in 40 countries,
- with 113 insurance companies,
- and 126 companies mostly,operating in the financial an real estate sectors.
Generali Gestion Santé:
- is #2 in France with 400.000 subscribers.
|
| | | about the service | Functions:
- automatic Call Qualification to the correct operator,
- consultation of the repayments available 24/7,
- breakdown of the guarantee ,available for certain contracts 24/7.
Key figures and results:
- 1000 calls per day on average,
- 75% users enthusiastic about natural language,
- immediate allocation of the calls,
- 10% of the calls are dealt with automatically,
- rate of calls taken improved to an average of 85%,
- reduced call times thanks to efficient call qualification,
- enhanced operational functioning of the platform.
|
|
For more information on Generali , please visit www.generali.com.
For more information on Orange Business Services, please visit www.orange-business.com. |
teliSpeech platform: open, simpler and faster |
| | Deploying speech isn’t easy, it is still complex and timely to both develop and tune your applications. In addition to this it is costly and difficult to deploy across different countries and regions. The products on the market are vendor dependant, meaning their application tools are not freely available. How then can a company deploy the best speech application? |
Telisma has taken a major step forward to remove these barriers, by adding new functionalities to the teliSpeech platform. Taking a comprehensive approach telisma has implemented a structured platform to address these issues that its partners come up against. The concept of the teliSpeech platform means that it can be built upon; third parties can use the individual products openly. It consists of three principal elements:
| - teliSpeech Framework
- teliSpeech Languages
- teliSpeech Recognition Engine.
|
teliSpeech framework |
The framework is built up of a range of Tool Kits and Development Kits which will progressively all be integrated in the Eclipse development environment; from acoustic tuning, creation of language variants or “Dialects”, creation of complete languages to intelligent call routing.
The Natural Language Took Kit (Natural Language TK)
The Natural Language TK is a core element of the teliSpeech framework. The main applications that are currently using Natural Language are customer-care and self-service. A good example of this would be a “How May I Help You?” welcome prompt.
| | | what is your business problem? | | As the market demand moves towards intelligent call routing for call centres, the demand for Natural Language applications is ever-increasing. | | how can telisma help you solve it? | With the Natural Language TK it is ideal to use the largest text corpus possible as to represents the typical interactions between a customer and a call-centre agent.
The Natural Language TK enables you to generate the SLM which is used by the teliSpeech Recognition Engine to output the textual representations of any spoken input. The output can then be parsed to spot those particular keywords of interest.
The call is categorized and routed to the appropriate agent or to another speech-enabled application dealing with that particular request. | | what are the business benefits? | | The software vendors and system integrators that have developed their own modules to deal with the semantic post-processing of a raw recognition result in various domains (finance, healthcare, insurance...) can now work with the teliSpeech platform. |
|
The Tuning Tool Kit (Tuning TK)
The Tuning TK is part of the teliSpeech framework. It helps you enhance the quality of your services, by putting you in control of tuning your speech application. Telisma can help you leverage your skills, make you independent of the speech vendor, and put you into control over your application.
| | | what is your business problem? | | Need to continuously improve the level of performance of your application? Feeling like you are too dependant on your speech vendor? | | how can telisma help you solve it? | The Tuning TK allows you improve the Quality of Service (QoS) of your speech applications.
Conduct an in-depth audit of the ASR performances for your applications and, if need be, improve your grammars accordingly, from the syntactic level right down to the phonetic and even the acoustic level.
Specific training is available. | | what are the business benefits? | Gain the appropriate skill sets in-house to deliver and deploy your speech applications without having to use expensive professional services resources from the speech vendors.
Gains the highest possible level of control on all stages of the speech applications lifecycle, from the design to the implementation and tuning phases. |
|
The Dialect Development Kit (Dialect DK) As part of teliSpeech framework, the Dialect DK allows you to fine tune the teliSpeech Recognition grammars or even adapt a teliSpeech language for any application to deal with strong regional accents and dialects.
| | | what is your business problem? | Are your end-users being understood? Need to deal with a strong regional accent problems, or to implement your voice service in a country using a variant of English, French, Spanish, German, Hindi, etc?
Take into account all your end users accents and improve your customer satisfaction. | | how can telisma help you solve it? | The Dialect DK allows you to create your own dialects by adapting an existing language pack or voice database to derive another language.
Specific training is available.
| | what are the business benefits? | The Dialect DK has a simple, easy-to-use interface.
Achieve autonomy, higher recognition rates and break through performance.
Have a complete overview of the application life cycle. |
|
The Language Development Kit (Language DK) telisma wants to make speech openness a reality worldwide, let everyone benefit from its innovative teliSpeech platform. Telisma wants its partners to share its open vision of speech and build their own languages.
| | | what is your business problem? | | Is your company or subsidiary based in a country where ASR is not available? Or you are looking to develop your own business in new territories? | | how can telisma help you solve it? | The Language DK is a complete methodology and set of tools that enable you to build your own language quickly and efficiently.
The LDK requires specific training. | | what are the business benefits? | Build your very own Intellectual Property using a complete set of tools and processes from telisma.
Use of an extremely user-friendly Eclipse-based Graphical User Interface.
Building your language is quick and simple, no need to be a speech expert! |
|
teliSpeech languages |
Telisma expands its teliSpeech Language catalogue and now offers Castillan Spanish for Spain, Dutch, Indian English and Hindi in addition to the performance improvements that have been made on the other languages already on offer.
Each of these new teliSpeech languages have been developed using the Language DK. teliSpeech Recognition Engine |
teliSpeech Recognition Engine is telisma’s speech recognition software. It is continuously updated and enhanced to take into account the new advancements on the market. Telisma has added new Operating Systems and Compilers to the current teliSpeech version. It is supported on the following platforms: | operating system | | Windows 2003 with Service Pack 1 | | Linux Enterprise 4.0 |
| | | compiler | | Microsoft Visual C++ 6.x or 8.x | | GCC 3.4.3 + binutils 2.15 |
|
|
nabaztag/tag - A hoard of users adopts the smart rabbit and speaks to it thanks to teliSpeech |
| |  |
Say hello to the Nabaztag/tag, the new generation of the intelligent and highly fashionable internet connected WIFI-bunny!
Speak to your rabbit thanks to teliSpeech
Hidden behind the rabbits belly-button is teliSpeech which allows the rabbits to recognise the various voice commands associated to the main functions; radio; weather, traffic in English, French and more languages.
So, by looking at him straight in the eyes and talking to this belly-button you can ask your favourite pet what the weather will be like on Sunday, who has sent you an email, ask him to connect to your favourite radio station, to read you the stock market reports, or the latest news headlines… He is remarkably obedient!
Technically, the speech recognition is performed by teliSpeech on the Nabaztag networked server; the rabbits simply act as terminals. New services can therefore be added at any time, without the need for any upgrade to the rabbit’s hardware! Among the other technologies used in the rabbits is the TTS from Acapela.
Thousands of rabbits are already connected to the Nabaztag networked community interacting with teliSpeech. After their success in the UK, USA and France, the Nabaztags will soon be up for adoption in other European countries. The rabbits are hopping their way to you…
Adopt a smart rabbit: www.nabaztag.com. |
|
calling the hospital gets more pleasant and efficient with Jet Multimedia and teliSpeech |
| | Jet Multimedia, a leading hosting provider in France for IVR services has commercialized a really innovative, highly performant, contact center health packs aimed at health centers such as hospitals. |
The pack consists of a speech enabled IVR solution, an easy-to-memorize phone number, and an administration extranet. A ready-to-use, complete, solution (Software as Service business model), the pack provides hospitals and clinics with a state of the art automated contact center, with no initial investment. With over one hundred hospitals and clinics already deploying this solution, it is proving it to be a real success in this sector! It has a huge impact on the callers’ satisfaction and frees up time from the highly skilled personnel.
In order to strengthen its leadership on this market, Jet Multimedia has gone one step further by integrating teliSpeech into its health pack. Users now simply have to say the name of the person they want to reach, and the call will be automatically connected. For hospitals and other health centers, the new speech-enabled contact center has proven to considerably improve client satisfaction, and to reduce holding-times and management costs. Moreover, it necessitates ZERO investment.
For more information, please visit www.jetmultimedia.fr |
training: new courses and certification levels to enhance your speech expertise |
| | New courses have been added to the teliSpeech training catalogue:
| - Master teliSpeech integration through MRCP or native API with the Integration Workshops program: 1 day
- Have an overview of the voiceXML, SRGS and MRCP standards with the Speech Technology Standards program: 1 day
- Manage natural speech recognition with teliSpeech with the Statistical Language Models program: 2 days
- Optimise your system’s performance using the teliSpeech tuning toolkit with the Acoustic Tuning Training program: 3 days
|
These courses will also be available through the teliSpeechUniversity, open to all professionals wishing to develop new skills and meet other actors of the speech ecosystem. Interaction between participants and sharing of experience will be at the core of this program.
Become a teliSpeech trained partner! Knowledgeable employees are a key element to the success of the speech application. Show your customers how you master the teliSpeech platform. Once two of your experts have successfully undergone the required telisma training courses, your company will qualify as a ‘trained partner’ with telisma and will be qualified to use the corresponding trained partner logo. The different logos represent the different levels of certification. There are three levels of certification; Bronze, Silver and Gold, depending on what type of training courses you undergo. What type of trained partner will you be?
For more information on teliSpeech training catalogue or trained partner program, please contact training@telisma.com or visit http://www.telisma.com/training.html | |
|
CII Industrielle and telisma join forces to make new generation speech services mainstream |
| | Created in 1990, CII Industrielle develops proprietary telecom applications such as: IVR, voice mail, information servers, unified messaging systems, auto-attendants, tele-payments systems, CTI, IPBX and VoIP. Its main references are in transport, call centres and CRM.
|
A pioneer in developing speech applications Philippe Jauneau, CII CEO says: “I’m glad to see that the market is finally ready to make speech recognition available to the mass market. The technology is mature and cost-efficient, and clients are now openly demanding this technology: the time has come to make it widely available”.
CII chose telisma as its exclusive speech recognition partner after completing comparative tests with other leading technologies both for the performance of its technology and for the quality of its commercial and support services. “The proximity and reactivity of telisma’s teams were both important criteria for our choice” confirms Philippe Jauneau. In order to benefit from a transfer of expertise from telisma’s experts and to gain the best from the technology, the CII team underwent a complete training program including integration, OA&M and Grammar Development sessions.
Following the signature of a reselling agreement, teliSpeech has been integrated on the CII Industrielle AS-VOX G4 IVR platform. Name and surname recognition functionalities will soon be offered by default to all customers, whilst more advanced recognition features including natural speech will continue to be offered on demand. “The CII Industrielle platform is very homogeneous and reliable. It’s one of the more complete on the French market in terms of functionalities and the fact that it directly integrates the ASR and TTS modules into its IPBX makes it very unique” says Laurent Balaine, telisma’s CEO. “We’re very pleased to have been chosen as CII’s exclusive speech partner;together we will boost the speech market with next generation solutions”
First deployment coming soon
CII Industrielle is currently developing speech-enabled timetable information service for a bus transport network which integrates telisma’s speech command.
For more information, please visit www.cii-industrielle.fr |
|
if you do not wish to receive information from telisma via email,
please reply to this email with UNSUBSCRIBE in the subject line
|
|
