Enelai building blocks / Libraries

Hundreds of enelai building blocks accelerate the development & optimization of your application by providing industry-specific language understanding & semantic interpretation models, conversational & users’ behavioral patterns, as well as prebuilt self-service dialogues to use as templates to modify & design your own. Created from the data collected in 2 decades of deploying real-world applications (and constantly adding new), enelai combines the benefits of machine learning with real-world deployment know-how.
Simply put, these libraries are >80% of the industry-specific Semantic Language Models both for speech & for text Natural Language Understanding, and more than 50% of the entire effort required to design, deploy & optimize a project.

Pre-Built, industry-specific Intent Libraries

Language agnostic, industry-specific Intent Libraries and self-service dialogues streamline & accelerate the implementation of each AI-powered Virtual Assistant optimized for highest User Acceptance and results. The industry-specific libraries cover 650+ Telecom Intents, 430+ Banking Intents as well as general Customer Service Intents from various industries, paired with the essential disambiguation dialogues where needed.

Many of these Intents are also combined with industry-specific, multi-turn self-service dialogues, encompassing best practices from real-world deployments and are optimized for high performance. Each sample is used as a working prototype to be modified and adapted as needed enabling rapid implementation and deployment of your application with self-services of high business value.

Telecom

Dialogues cover self-service frequently-asked Intents / User Tasks, such as Βill Details, Payments, Select & Purchase Add-on services, Activate 4.5G, Troubleshoot Internet, Change Tariff plan, Promise to Pay, Technician Appointment, Change Number, Retrieve PIN/PUK, Network Coverage, Campaigns, White Pages/Yellow Pages directory services, and others.

Financial/Banking

Dialogues for self-service of User Tasks such as Card Statements, New Card Selection & Application, Card Activations, Retrieve/Reset PIN and Web Banking codes, Make Payments, Transfer Money, Troubleshoot Internet Banking, Locate ATM/Branches, Promise to Pay, Change/Update personal details, etc.

Pre-Built, authentication methods/factors

Integrated authentication methods for parameter-based configuration without coding: simply select what (if any) Authentication Strength is required for each Intent (None, Light, Medium, Strong), and then which Factors are to be employed for each strength, in each deployed channel.

Configure the Authentication Factors applicable in your operations

Select the factors to be used for each Authentication Strength (Light, Medium, Strong)

client

system
or agent

Voice Biometrics

With biometric technology, customers are authenticated naturally with the sound of their voice, their face or even the way they interact with the device. Authenticate your customers in the background while they make their requests (to the system or to the agents), without interrupting their requests and without the need for them to remember long numbers & codes as additional authentication factors.

What type of Biometrics are available

Authenticate your customers with a passphrase such as “my voice is my password”. 

Authenticate your customers when they make their requests (to IVR or to Agents) in the background, without interrupting their requests.

Give your customers the choice of Facial Recognition on their mobile phone or laptop.

Continuous authentication to ensure the customer with the log in is really the customer.

A form of behavioral biometrics that can identify fraudulent activity in real-time based on choice of words, and patterns of speech or writing, during an interaction with a human or virtual assistant.

With behavioral biometrics we are able to authenticate and protect each user’s personal identity based on patterns in their behavior: how a person types, uses a mouse, holds their smartphone or even the number of pauses in their speech. These patterns create a personalized profile which is compared to each user’s real-time behavior when interacting with the interface, ensuring that no fraudulent activity is taking place.

AI-powered technologies layered together form a profile to verify legitimate users and flag fraudulent activity including spoofing attacks using synthetic speech, recording attacks, mimics, network, devices and location changes.

Contact us
for more info and case studies