Chatbots Testing Automated testing for Chatbots Jan, 2018.

Slides:



Advertisements
Similar presentations
Test Automation Success: Choosing the Right People & Process
Advertisements

Achieving Success With Service Oriented Architecture Derek Ireland 17th March, 2005.
Accelerate Business Success With CRM CRM Interoperability.
What is Business Intelligence? Business intelligence (BI) –Range of applications, practices, and technologies for the extraction, translation, integration,
Powering your CTS Investment with Diagnostics and Resolutions Paul McCloskey.
Spoken Dialog Systems and Voice XML Lecturer: Prof. Esther Levin.
Intelligent Performance Management Empowering Your Enterprise Duane E. Presti, CEO PARIS Technologies, Inc.
©Copyright Artificial Solutions 2015 Artificial Solutions & the Teneo Platform Making Technology Think September 2015.
0 What Does SIP Bring to Your Customer Experience ? Extend VoIP and IP Contact Center values through support of SIP o Media and location independent support.
Sales POV: Digital Customer Services Platform Presented by Michael Bommer.
Reinventing Customer Experiences
Why industry cares about nlp for tamil?
AuraPortal Cloud Helps Empower Organizations to Organize and Control Their Business Processes via Applications on the Microsoft Azure Cloud Platform MICROSOFT.
Case study 05 Ana-maria casian, Natalija mitrovic, sam tecle, clyde pietro & ryan davies.
Transforming business
Topics Question answering at Bing
Employee Self-Service
Intelligent Agent Solution
Let’s talk about Conversation Design
Primal and Microsoft Azure Deliver Personalized Content, Intelligence, and Analytics That Match Your Content to the Interests of Your Audience MICROSOFT.
Introduction to Azure Bot Framework
Customer Relationship Management (CRM)
Session code.
Build a website like Amazon/Ebay/Etsy/Airbnb - experiences and advice.
Asad Khan Managing Director Health & Public Service.
Chapter 18 MobileApp Design
Helping Millions of digital banking customers be…
© 2016 Global Market Insights, Inc. USA. All Rights Reserved Fuel Cell Market size worth $25.5bn by 2024Low Power Wide Area Network.
Language Understanding Intelligent Service and Microsoft Azure Enable Rover, PLEX.AI’s Artificial Intelligence-Powered Virtual Insurance Advisor MICROSOFT.
Get Real Value and Insights from Your Data: Biin Solutions Provides Predictive Analytics, IoT, and Business Intelligence with Microsoft Azure Power MICROSOFT.
Take Control of Insurance Product Management: Build, Test, and Launch Any Product Globally 10x Faster, 10x More Cheaply with INSTANDA on Azure Partner.
© 2016 Global Market Insights, Inc. USA. All Rights Reserved Chatbot Market to reach $1.34bn by 2024: Global Market Insights Inc.
OpenWorld 2018 How to Create Chatbots with OMCe
Developing Innovative Unified Communications Applications
CTI Contact Center For CustomerSoft ESP
Chatbots for Dummies José 10/11/2018 Immersion
INTELLIGENT (SMART) E-COMMERCE
Continuous Automated Chatbot Testing
Be Better: Achieve Customer Service Excellence and Create a Lean RMA and Returns Process with Renewity RMA and the Power of Microsoft Azure MICROSOFT AZURE.
A STEP-BY-STEP GUIDE FOR ANDROID MIGRATION
Service Desk Automation BOT
Voice Analytics on Microsoft Azure Allows Various Customers to Get the Most Out of Conversations with Clients Through Efficient Content Analysis MICROSOFT.
José Mendes Chatbots for Dummies.
Chatbots Best Practices & Design Patterns…
11/23/2018 8:30 AM BRK3037 BRK3037: Dive deep on building apps and services with the Office 365 Communications Platform David Newman Senior Program Manager.
Speech Capture, Transcription and Analysis App
Collaborative Business Solutions
One-Stop Shop Manages All Technical Vendor Data and Documentation and is Globally Deployed Using Microsoft Azure to Support Asset Owners/Operators MICROSOFT.
Appcelerator Arrow: Build APIs in Minutes. Connect to Any Data Source
Core Platform The base of EmpFinesse™ Suite.
Media365 Portal by Ctrl365 is Powered by Azure and Enables Easy and Seamless Dissemination of Video for Enhanced B2C and B2B Communication MICROSOFT AZURE.
E-Commerce Chatbot by Coviam Technologies
Top Reasons to Choose Android Today. Over the years the Android OS has progressed largely by acquiring major percent of global market share. A number.
Top Reasons to Choose Android Today. Over the years the Android OS has progressed largely by acquiring major percent of global market share. A number.
Technical Capabilities
Chatbots for Dummies José 12/05/2018 Immersion
Facebook Enterprise Point of View.
Why IBM Watson.
Technical Capabilities
Agile testing for web API with Postman
Single Cell’s Progenitor Powered by Microsoft Azure Improves Organisational Efficiency with Strategic Procurement, Contract Management, and Analytics MICROSOFT.
Open Source Tool Based Automation solution with Continuous Integration and end to end BDD Implementation Arun Krishnan - Automation Manager Maria Afzal-
SAP Hybris Service Cloud
Guided Research: Intelligent Contextual Task Support for Mails
Building and Integrating a Chatbot in 30 minutes
IBM AI for Customer Care
A STORY ON CONVERSATIONAL CHATBOT SUBHAJIT SENGUPTA & SHARIQ SHAIKH
Pitch Deck.
TN19-TCI: Integration and API management using TIBCO Cloud™ Integration
A Cryptographic Token for Building the Smartest AI Entity by The Crowd
Presentation transcript:

Chatbots Testing Automated testing for Chatbots Jan, 2018

Rise of Chatbots The new bot trend is due to two main factors: artificial intelligence and the rise of messaging applications. “People are now spending more time in messaging apps than in social media and this is a huge turning point.” “Messaging apps are the platforms of the future and bots will provide users with all sorts of services.”  - Business Insider Rise of Messaging Apps

Bot Ecosystem Key Components in the bot architecture are Chat apps, NLP layer & Enterprise Backend Enterprise database to seed Chatbot with the information to respond to customer questions 3 Data store captures data on users’ activities, hits & misses in Chatbot responses 4 Chat apps like Facebook Messenger, Slack, etc. are the interface through which users access and interact with your bot 2 Natural Language Processing translates users’ free form questions into information that can be mapped in real-time to formulate appropriate responses 1

Trending Chatbot Techstack Chatbot platforms Chat messengers Chatbot Platforms backed by Tech Giants Api.ai (Google) Wit.ai (Facebook) LUIS (Microsoft) Watson (IBM) Lex (Amazon) Non programming Chat bot platforms Chatfuel ManyChat Octane Ai Massively Motion.ai.

Testing tenets in Chatbots Testing zones Key testing areas in the bot architecture are highlighted in the below schema DB Tests API Tests Analytics pattern checks Intents & Utterance tests Chatapp Voice, Video and Text validations Natural Language Understanding (NLU) Tests Bot Interoperability testing with messengers

Chatbots Testing Approach 1 Voice & NLP text Testing API Testing UX Testing 2 3 Provide multi accent user intent in the form of Voice/ text to Chatbot and verify its capabilities to process the required ‘Intents’ and respond with the appropriate ‘Utterances’ Use Automated Text to Speech and Natural Language Generators to create combinations of user input & validate the correctness of the Bot responses Understand API Descriptions and endpoints between the Bot UI, NLP Layer and the Backend All combinations of Straight-line cases, boundary cases, null input, Idempotent verbs can be automated and tested at the API level Use Built in assertions in API testing tools like Postman, SoapUI to verify the responses from various layers Assess 3rd Party API Dependencies For Vulnerabilities and perform security testing Achieve Virtualization of backend for faster testing Perform Interoperability testing of Bot with multiple chat messengers and mobile devices(OS version, Device manufactures, form factors) Perform User experience and usability testing of the chat application Measure Mean time/loops to complete transaction and the number of chats to achieve end result Perform NFR testing in targeted mobile and tablets

DAVO - Test Automation framework for Chatbots DAVO Framework 1-Turn Chat Multi- turn chat Automation Libraries Lex API Endpoint Chat UI Utterance Generator Utterances for Intents Utterances with slots DAVO Chat UI Input Text … Enterprise Services

Advantages of DAVO Framework Cross Platform Single Script supports multi platforms like Amazon, FB, Slack Crowd Enabled 360° Test Automation E2E framework to test all layers of Chat, voice, API and Backend Seamless Integration with Crowd platform for exploratory testing Assured CX Voice/Chat app tested with bulk natural user utterances for a successful production Day 1 QA History Shorter Cycle time Maintains HTML Based reports for all chat utterances High reproducibility and Accelerated Feedback to developers for skill tweaks

Lex Chatbot implementation for a Large American Insurance Provider Client is a North America based Insurance company who is one of the US largest insurers of vehicles, homes and small businesses and provide a wide range of other insurance and financial services products with annual revenue of $19B and have 20,000+ employees. Business Problem – The project objective is for Chatbot to deliver canned comments such as initiating, Idle time & ending the chat, guidance to agents on some transactions & answers to FAQ questions there by enabling a quicker response to agent’s inquiries. Customer Situation Farmers is looking at improving the agent experience by reducing hold time for agent servicing through chat and reduce overall costs of call center operation.  Cognizant team analyzed the call center data and determined reasons for agent calls/conversations. Cognizant proposed Amazon Lex as an option to automate the chat conversations. Amazon Lex is a service for building conversational interfaces using voice and text. Lex‘s automatic speech recognition (ASR) and natural language processing/understanding (NLU) capabilities allows you to create self-service experiences. Business Benefits Reduction in call and chat volume to the Service Center Solution Details Intent Keywords are provided as Inputs ( FAQ’s, status of Document uploaded / Policy & Claims). User sessions are to be short and exact the need. When the Chatbot doesn’t able to resolve the customer's issue, Amazon Lex is able to pass on the full context of the conversation to an available CSR representative which keeps the customer from wasting time repeating answers to questions and lets the CSR representative focus 100% of their time on solving the problem, which increases the odds the customer is going complete the contact feeling positive about the experience. Effort Saving up to 30% where there is no need for CSR to be available all the time for Customer initial queries. Different way of responses are been designed and tested like following:- No response — for instance, if someone is idle for sometime. An invalid response — the wrong message entirely not relevant to the chat. A valid response with invalid data —  the correct container of query but with an exclusive word. A valid response — any of a number of variations. Bot only Chat  95% time savings for CSR Bot + CSR Chat  5% time savings for CSR CSR only Chat  0.05% time savings for CSR Client Name: Farmers Insurance NPS: A metric to measure customer feedback GWP: Total premium written and assumed by an insurer before deductions Combined Ratio: (Incurred Losses + Expenses)/Earned Premium