Creating Smarter Conversational Experiences with Infinity Botzer on AWS

By Austin Sequeira, Head of Cloud Innovation – LTIMindtree
By Lewis Tang, Sr. Solutions Architect – AWS
By Francois van Rensburg Sr. Solutions Architect – AWS

LTIMindtree

Artificial intelligence (AI) is accelerating digital transformations, changing the very facet of customer-driven businesses. With conversational AI bots gaining traction over rule-based conventional chatbots, many organizations are investing in intelligent virtual assistants for improved customer service.

Traditional or click-based chatbots are relatively simple and provide a predetermined set of basic information or responses for users to choose from.

AI bots, on the other hand, are more sophisticated and use artificial intelligence to understand and interpret user inputs. They are best suited for enterprises with enormous data, ensure faster response time, and are available 24/7.

Using machine learning (ML) algorithms and natural language processing (NLP), conversational AI bots are able to understand the context and objective of a question. This helps them learn and grow from interactions and gain better insights into user preferences and habits.

With minimal training, these AI-powered assistants are able to respond to user queries generating impactful user experiences.

In this post, we will look at how enterprises can build and deploy bots within minutes using Infinity Botzer from LTIMindtree. We’ll also help you understand how this platform can facilitate comprehensive bot lifecycle management for businesses to become more efficient, responsive, and customer-centric.

LTIMindtree is an AWS Premier Tier Services Partner and Managed Service Provider (MSP) that enables enterprises across industries to reimagine business models, accelerate innovation, and maximize growth by harnessing digital technologies.

Challenges and Key Benefits of Infinity Botzer

Keeping pace with the dynamics of customer service is complex and includes many challenges:

• Understanding, prioritizing, and exceeding customer expectations and needs.
• Effecting personalized services that match specific needs and preferences.
• Mitigating repetitive interactions and long wait time.
• Providing continuous business support at scale.
• Enabling high-speed query resolution.

The rich experience and in-depth understanding of consultants in the areas of data science, machine learning, NLP, and multiple chatbot technologies has enabled LTIMindtree to build a smart conversational AI platform that provides high levels of flexibility and accuracy.

Infinity Botzer is a software-as-a-service (SaaS)-based conversational AI platform that helps build digital assistants and automate conversational workflows. It provides accelerated, engaging, and personalized user experiences.

The Infinity Botzer chatbot solution provides self-service capabilities and automation, resulting in improved productivity, optimized operational costs and faster time to market.

With Infinity Botzer, enterprises can build and deploy bots at a fast pace and create intuitive conversational flows for inclusive end-user experiences and optimized customer service operations.

Infinity Botzer is helping customers:

• Achieve impactful user engagements.
• Gain higher customer satisfaction.
• Avail instant round-the-clock solutions.
• Improve end-user productivity with faster resolutions.
• Enhance support agent efficiency.
• Scale to handle larger user volumes.

Solution Overview

Infinity Botzer’s core competencies are built using Amazon Lex, Amazon Polly, and Amazon Simple Storage Service (Amazon S3). The NLP capabilities of Amazon Lex are leveraged to build, deploy, and monitor scalable, easy-to-use AI conversational bots. Amazon Polly, powered by its deep learning engine, helps in synthesizing text to a human-like voice that has improved speech quality.

Static content like images, graphs, and charts generated alongside chatbot responses are stored on Amazon S3, and AWS Lambda is used to compute and pull out relevant data from the stored S3 buckets as a response to end user’s queries when required.

Infinity Botzer is a do-it-yourself (DIY) solution that aids business users with nearly no code experience to build and deploy enterprise bots effortlessly. These ready-to-deploy bots are capable of understanding customer interactions, as well as their behavior patterns, to align with customer needs to make businesses more responsive, flexible, and service-oriented.

Infinity Botzer’s cognitive engine effectively auto-trains bots with ease, helping create smart human-like conversations.

Key features of Infinity Botzer include:

• Smart business workflows: Craft dynamic conversational workflows with industry-specific and pre-built templates, conditional routing capabilities, and a rich user interface-based preview to generate engaging and personalized user experiences.
• Real-time insights: Evaluate bot performance and success metrics through advanced bot analytics for increased operational efficiency.
• Personalization: Customized bot settings are tailored to enterprise needs. Example: on-the-go bot themes, multi-lingual, multi-channel, and voice-based hands-free experiences.
• Cognitive engine: Infinity Botzer’s ability to gather and leverage user behavior and preferences as a knowledge base via its AI/ML capability enables continuous bot learning and feedback, leading to increased accuracy and enhanced user experiences.
• Integration: Pre-engineered design components and API-based services enable seamless integration with enterprise systems.

Infinity Botzer platform’s frontend consists of two key building blocks:

• Botzer management portal: Where bot designers can build and configure chatbot workflows, as well as define the channels on which chatbots will be deployed.
• Chatbot interface: Through which end users or visitors will interact with the chatbot and live agents for human assistance.

Additionally, Infinity Botzer comes with an in-built Live Agent portal which allows end users to seamlessly connect with human agents for  supplementary assistance during chatbot conversations. The human agents are categorized per skills or teams in the Live Agent portal.

The chatbot workflows and conversation history can be configured to connect with human agents per their skillset, so user queries are addressed with the least turnaround time to provide the best possible customer experience.

Supported user personas include:

• Admin: Controls and manages overall platform configurations like training chatbot, designing workflows, configuring external integrations, onboarding new users, and providing platform access while modifying existing access and revoking user access.
• End users: Visitors who interact with chatbots and live agents to seek required information and get their queries resolved.

How it Works

Infinity Botzer is built on a microservices architecture that helps easy scaling and faster development, thus accelerating time to market for new features. The platform’s microservices can communicate with each other using REST APIs to reduce latency and achieve faster responses between the microservice components.

Figure 1 – Architecture diagram.

Infinity Botzer’s network architecture consists of:

• Amazon Route 53 which is a reliable and cost-effective domain name service (DNS) to route end users to Infinity Botzer.
• Private and public subnets in an Amazon Virtual Private Cloud (VPC).

All architectural components are in private subnet except the Application Load Balancer (ALB) and Linux bastion host. The ALB automatically distributes incoming traffic from client systems, thus increasing the availability of the application. The Linux bastion host acts as a gateway between private subnet and the internet, allowing client machines to connect to the remote server.

Amazon Elastic Kubernetes Service (Amazon EKS) hosts the platform’s front-end user interface (UI) layer, and helps run and scale on the AWS cloud for high-availability, reliability, and security of the frontend applications. Amazon EKS integrates with ALB for load distribution to ensure application availability at any given point in time. Amazon EKS, along with frontend applications and backend service APIs, are structured as PODs to manage and scale application deployment process.

Kubernetes controller helps manage and automatically scale the number of required instances of an application based on demand. For example, if there is a sudden spike in traffic detected for chatbot conversations, the controller can increase the number of application instances to handle the load and ensure chatbot responses are not delayed and end user experience is not hampered.

The controller also monitors the health of PODs and restarts them, if necessary. This helps minimize downtime and maintain the overall health of the application.

Amazon ElastiCache for Redis is a redundant cluster with encryption enabled for data security and protection. Application servers use Redis cache to handle large amounts of real-time data, store session information, and cache API responses for faster performance. Its in-memory data storage capabilities help support highly responsive database constructs.

Amazon Relational Database Service (Amazon RDS) for PostgreSQL is used as the persistence layer that stores user data and is deployed in the private subnet with specific ports open to application security groups. It has backup, detailed monitoring with alerts enabled, and maintains data privacy through encryption algorithms.

PostgreSQL is used to store all chatbot configurations, metadata, and audit reports. Amazon S3 stores the multimedia related chatbot data such as themes CSS, JavaScript, images, videos, and bulk operation files data. Elasticsearch is used as a secondary storage that helps to speed up data search and retrieval.

All information related to platform configurations such as chatbot workflows, NLP training data, API configurations, and reporting history is securely stored in a database along with encrypted customer session data. Access to the database is controlled by AWS Secret Manager-based authentication within HashiCorp Vault.

Jenkins is integrated with Amazon Elastic Compute Cloud (Amazon EC2) for continuous integration and continuous deployment (CI/CD) that helps streamline the deployment process, improve scalability, and provide monitoring capabilities in real time.

In addition to architectural components, Infinity Botzer consumes capabilities of the following AWS services to enhance the features and functionality of the platform.

Amazon Lex

Infinity Botzer is integrated with Amazon Lex for NLP capabilities, in addition to custom algorithms for improved intent detection and better accuracy.

The Amazon Lex-supported interface within Botzer management portal allows you to build, deploy, and operate simple and complex bots leveraging natural language understanding (NLU) technologies and automatic speech recognition (ASR) capabilities. This helps build highly engaging conversational workflows.

Amazon CloudWatch

Infinity Botzer uses Amazon CloudWatch for monitoring and analysis of real-time logs, metrics, and event data. CloudWatch logs help monitor Infinity Botzer’s resources on AWS and trigger alerts extensively to ensure applications are performing efficiently. It also helps speed up debugging of applications in case of any issues and reduce overall mean time to resolution.

Amazon Polly

Infinity Botzer is integrated with Amazon Polly for voice recognition capability and to synthesize natural-sounding human speech. With dozens of lifelike voices across a broad set of languages, Amazon Polly accelerates Infinity Botzer’s text-to-speech and speech-to-text conversion capabilities and responds to users’ voice-based interactions and queries.

Amazon SES

Infinity Botzer uses Amazon Simple Email Service (SES), an SMS service used to send notifications to end users on variety of chatbot use cases such as service status, request acknowledgments, confirmation emails and much more.

AWS IAM

Infinity Botzer uses AWS Identity and Access Management (IAM) to centrally manage access to different resources and services using fine-grained permissions. IAM is also used to control access privileges to refine permissions across AWS infrastructure landscape.

Account Onboarding

Users can create an Infinity Botzer trial account for 30 days of free access:

• Visit the Infinity Botzer admin page.
• Click on “Sign Up Now.”
• Enter the required details on the account registration page and click “Next.” You’ll receive a verification code to the email ID entered in account registration page.
• Enter the verification code, set a password for your trial account, and submit.
• On successful email verification and password creation, you will be redirected to the admin page.
• Log in to your Infinity Botzer trial account using registered email and new password.

Here’s how to create a business account with Infinity Botzer:

• Reach out to the Infinity Botzer team by writing to infinitycloud@lntinfotech.com.
• Receive login (email and temporary password) credentials from Infinity Botzer.
• Visit the admin page.
• Enter your email and temporary password received from Infinity Botzer.
• You’ll be prompted to reset your password when you log in for the first time.
• On successful password reset, you’ll be redirected to the admin page.
• Log in to your Infinity Botzer business account using registered email and new password.

Summary

In this post, we saw how Infinity Botzer helps build, deploy, and manage enterprise-grade chatbots rapidly and with ease. The solution’s chatbot-driven self-service and automation capacity enables enterprises to handle larger volume of requests and provides highly responsive and cost-effective business solutions, improving customer experiences.

Infinity Botzer facilitates easy chatbot integration with multiple channels and has a built-in live agent console that allows seamless transition from a bot to a human agent when required.

With its NLP and AI-based offerings, Infinity Botzer is transforming customer self-service into an evolved and future-ready industry.

.

.

LTIMindtree – AWS Partner Spotlight

LTIMindtree is an AWS Premier Tier Services Partner and MSP that enables enterprises across industries to reimagine business models, accelerate innovation, and maximize growth by harnessing digital technologies.

Contact LTIMindtree | Partner Overview