We recently launched Angie (internally) for some customers. Angie is a powerful chatbot application that can help people in a company to apply for leave, allow the managers to approve these leave requests right within Slack. It is probably the most advanced Slack Chatbot for HR we’ve ever made after UnderstandBetter for Slack.
While this is a cloud-based implementation, the infrastructure cost and other integrations with on-premise solutions are very limited. But many of our customers, who are using Angie are enterprise customers who require integrations with their systems that run inside their data centers.
This called for a new process of gathering the requirements from our customers and here are the 11 questions that we ask our customers to answer before we try and implement Angie bot framework for them.
- What business process are you trying to automate with chatbot?
- Who are the primary users of this chatbot?
- Will there be any hierarchy for approvals in the organization via the chatbot?
- What authentication method would you prefer? SSO, Microsoft AD or other oAuth providers?
- What are the top 5 features/functionalities that you expect in the chatbot for the first version?
- What software versions are you using internally with which the chatbot will be integrating to?
- Will this be a conversational chatbot or a guided chatbot? (we will be showing a demo of both the types of chatbots to the customer)
- Are there any other chatbots that you have already implemented?
- How deep of analytics would you require based on the users’ interaction with the chatbot?
- Will there be dynamic learning modules for the chatbot? Like learning from the internal KB documents and such.
- What other ethical considerations we should keep in mind for your organization while implementing this chatbot?
These questions have helped us get a clear picture of what the customer wants from a chatbot like Angie. Is there something you do that helps you understand more about the chatbot for your organization? If yes, hit us up on Twitter (@SkcriptHQ) and share it with us.