Altis Chatbot

In our previous blog on “The World of Bots” we discussed some use cases for Bots.  In this blog we will talk about a Bot that we created internally, the Altis Chatbot and some of the lessons we learnt in developing the Bot.

The Altis Chatbot is a resource and skill planning solution.  It has been designed to assist us to quickly and easily find team member details to match to project needs, including their contact details, types of projects they have worked on previously and their skills and experience.  We can also search by technology or toolset (e.g. Reporting > Power BI, or ETL > DataStage) and find team members who have those skills and details of our recent projects where the technology or toolset has been utilised.

The Altis Chatbot was developed by some of our Managed Services consultants, Warinda, Jin and Swathi as a proof of concept to address a real-world problem in Information Management.  In two months, the team built the chatbot including performing the following tasks:

  • Requirements gathering
  • Toolset evaluation
  • Building Data Models
  • Building the Bot
  • Testing
  • Documentation
  • Presentation to broader Altis team

Toolsets used

As mentioned the team evaluated a number of tools for developing the bot, but settled on the following:

  • Microsoft Azure Bot Framework
  • Microsoft SQL Server 2016 database
  • Skype as the front end

This combination of toolsets was selected for it’s smooth integration and ease of use. The Microsoft Azure Bot Framework supports a guided development process, but it wasn’t straightforward as anticipated (as with most new projects and technologies).

Data Sources

We trained our bot using the following internal datasets:

  • Altis Historical Project Register
  • Yammer – our internal collaboration tool
  • Replicon – our time recording system
  • Altis web site

Lessons learnt

During the development of the Altis Chatbot we learnt a number of lessons:

  • Conceiving the idea of the functionality we wanted the bot to have is one thing, but you need to make sure that you have the data available to deliver. In our case, we didn’t have level of detail for projects that we needed for this project, so we had to collect and collate this data from other data sources.

This is something we see our clients facing as well – they have an idea for a data or analytics solution, but the data isn’t being captured at all or is being captured but not in enough detail or no history is being kept.  In these cases, like with our project, we looked to see if we could source the data from other systems or start collecting and capturing the data that is needed.

  • While much of the functionality came from the Microsoft Azure Bot Frameworks as templates, we still needed some key development coding skills. In our case, we needed to do quite a bit of C# development to implement the final solution.

How to best leverage the Microsoft framework

By utilising the Azure Cortana suite, you are able to bring together Bots, Analytics, Power BI visualisations and a wealth of other powerful functionality.  With all this functionality, it is important to plan and develop a roadmap on how you might want to use the data so you can deliver to a plan.

Since developing the Altis Chatbot we have gone on to assist some of our clients develop their own bots, including developing a bot responding to common help desk/service desk queries – this bot essentially triages level 1 calls to allow for value add work by the team manning the help desk and call centres and even out peak periods for the team.

If you’d like to find out more about how Altis can help you build a bot for your organisation, get in touch at connect@altis.com.au or contact your local regional manager.

Join the conversation

Your email address will not be published. Required fields are marked *

Comments

Post has no comments.