In a previous blog, we explained how you can use ipushpull to make your data available on different platforms and to different workflows. In this blog we'll cover how to roll out a reference data service to chatbots, using FOW’s recently launched chatbot as an example. The process and concepts can be easily extended to other data sets.

Understanding end-user requirements

The first step to creating the best chatbot solution is to define how your data is to be used and who needs access to it. For example, does the end user need to search across a large data set to match specific criteria or do they need to be alerted when a data value changes? 

FOW's chatbot, for example, enables end users to interrogate a broad universe of reference data sets and get immediate responses from a central service. Previously access would be via a separate department in the customer’s organisation consuming large data files on a daily basis sent from external data providers via SFTP and using internal systems to somehow make the data available, often requiring multiple manual steps by the end user, relying on the enrichment of data in legacy systems or reaching out to other teams to obtain the required result. By using a chatbot the end user can now request the data directly from the data source within their preferred chat platform (for example Symphony, Microsoft Teams or Slack) and get an immediate response.

Data definition and handling

Once you've understood the end user requirements the next step is to agree on an appropriate model for your data. For real-time data, ipushpull supports Live Pages (which can be thought of as a range in a spreadsheet and have a maximum size of about 50,000 cells). For periodically updating larger data universes ipushpull supports Large Data Sets (which provides direct integration with MySQL, PostgreSQL and Oracle databases). You define the data model as a Schema within ipushpull which provides structure around your data, such that it can then easily then be exposed to your ipushpull chatbots, or other applications using easily configurable Views. 

FOW's Views were quick and easy to configure and included many different data types including numbers, dates, text, and web links, all relevant to their many use cases so that their end users would have the best possible UX. 

Data definition schema

Configuring your end user UX/UI 

Having set up your Views for different users, the next stage is to surface them in the end-user application - in this case, chat applications. You can configure your ipushpull chatbot to define the chatbot syntax and UX for each use case. A chatbot may be configured with a command to use multiple parameters which then carry out a time-based filter with a partial text search. And the beauty is that now you've defined and structured your data without using any code!

Walk through of the FOW chatbot running on Symphony

Your new UI may be focussed on on-demand chatbot access, but the structured data you've created can now be easily configured, for example, to trigger a notification if a particular condition is met or used to trigger a daily report such as market events due today. Live data can also be made available in Microsoft Excel or a web app, meaning the end user can see data updates as they occur. Your UI can also be branded to your preferred colours and can include help guides and different types of data delivery such as email or chat.

In just a matter of hours, FOW was able to work with ipushpull to configure new chatbots with a whole stable of powerful commands for their end users, combined with a set of pre-configured notifications, all delivered into a choice of end-user applications.         

Automating and Scaling up

Beyond your slick new end-user interface, ipushpull also offers a host of features for easy onboarding of new users and scale-up of existing customers.

For example, auto-enrolment makes the new user sign-up experience seamless while automated follow-up alerts into chat or email, which can be used to nurture new customers, leads or users through their early days of using your chatbot. Driving better user engagement ultimately means fast user onboarding and faster scale-up of your service.

Better User Oversight and Insight

Having onboarded your new users, your ipushpull service will monitor all activity with custom reporting and analytics providing all the information you need to really understand your end user's requirements and service them even better. A client activity dashboard lets you monitor exactly who's looking at what information and from which applications.

This has provided FOW with a new window into their end users' behaviour and requirements.  

Example usage analysis dashboard

In this blog, we've explored how using the ipushpull platform, and without using any code, FOW was able:

1. Increase accessibility and distribution of their data through a  host of new channels 
2. Build a foundation for fast onboarding and growth
3. Create a virtuous flywheel where deeper insight into their end users helps to improve the service

 If you are interested in finding out more why not listen to the podcast with Symphony and FOW.