Highlights
- Automation of commonly performed tasks in enterprise IT using a chat interface.
- ETA reduction from a few hours <5 minutes for most tasks.
- Pioneering practical application of a chat bot in enterprises.
- Eliminate errors in enterprise IT deployments and support service task executions.
Executive Summary
Indexnine helped evolve and expand on the idea of automation of devops and project
management tasks using an intuitive natural language chat interface.
The product aims to transform the experience and effectiveness of IT support personnel
performing daily tasks to support incoming requests. The product can be used directly by end
users to perform several if not all devops tasks ranging from deployments to backups to regular
day-to-day project management tasks.
- Integrations with JIRA, Ansible, Github and other tools.
- NLP based intent-purpose identification and processing with API.AI.
- Rich user interface widgets to display results in an engaging and intuitive way.
- Guided and stateless chatbot implementation ready for extreme scale without
compromising on user-context.
About the Customer
Digitamize is an enterprise focused digital innovation agency, bringing disruptive perspectives to achieve business outcomes through augmented intelligence.
Challenges
- Identify Intent-Entity combinations across multiple products based on natural language input.
- Create engaging user task-flows by enabling a step by step user input flow.
- Allow end-users to define and expand the task flows.
- Design for horizontal scale and avoid server session specific contexts.
Solution
We suggested and implemented a stateless chatbot architecture driven by a json-based language devised specifically for this purpose. The language supported if-then-else like statements and for-each like looping constructs to define user task flows within any intent-entity combination. User input processing was done using API.AI (now DialogFlow) to allow rich language input. To supplement input requirements, we built a set of widgets that allowed a guided input flow for the user. For reporting, we created a set of charting widgets and corresponding interfaces to backend data that described the data and widget as part of the chat bot’s json flow definition.
What the Customer Has to Say?
With our solution building expertise, we were able to support this pioneering achievement with the right combination of UX/UI, solution architecture and web development to build this end-to-end product beta in a very short time.
Result
- High performance, scalable digital platform to provide a self-service chat interface for IT support consumers.
- Eliminate errors by providing a guided flow to IT support engineers to implement automated workflow tasks.
- A combination of a beautiful looking UI and an expandable robust solution for building guided workflows without writing code.