Our bot development has created new opportunities for business. The Bots that we have developed have enable staff and stakeholders to interact with the business 24 hours a day, 7 days a week on an unprecedented level of scale, using natural language, both text and speech.
We integrate Microsoft's LUIS (Language Understanding Intelligent Service) to allow employees to use Natural Language search. For example letting them run price queries, stock queries, ordering, account status checking, contact updates and preference setting by emulating human to human interaction.
As well as LUIS, we use the Microsoft Bot Framework. This reduces the time needed for the development process, as it provides a rich and full feature to make interactions between bots and users much simpler.
It also provides access to the other Azure AI and Cognitive Services as well as the Cortana engine. Bots though need training, they cannot work out what to do by themselves. We use our Business Analysis skills to understand your business and train the Bots.
We've set out our approach below:
Proof of concept
The first step is to create a Proof-of-Concept that demonstrates the potential and limitation of a Bot to the stakeholders using the available data set. This helps stakeholders understand the concept and the technical approach.
Bot or not workshops
The second step is to help stakeholders understanding what is possible shaped a series of workshops. The goal of those workshops is to determine if consumers or the business (depending if it is a B2C or B2B bot) could generate enough bot friendly questions to justify the development of a bot. Then to map out if those questions were of enough value (to the consumer and business).
Groups of employees are given the task of generating questions assuming they could query specified data sources and functionality. We then work with them to determine if the structure and content of those questions were answer-able (easily) by a bot using the Microsoft Language Understanding Intelligent Service (LUIS).
The outcome gives us a business case and test questions for development.
From there, we focus on:
- Bot development
- LUIS integration
- LUIS training
- Bot deployment
- Security and encryption
During the build we factor in in "Refinement questions" to focus the results and an in-Bot feedback process, which provides structured feedback from the test user base. This feedback is structured and allow us to train LUIS further on the "utterances" and the "intent".
The bot is then run in Test mode for a number of weeks to help refine its understanding and "intelligence" before being release in Beta.