Data Semantics Innovation Lab has developed a chatbot to address the complex management structure, enabling multiple user access levels for a single user. The bot further specializes in identifying the nature of the query by processing massive amounts of information to deliver relevant answers swiftly.
What does that mean?
In a typical organization, an employee can be a manager in one portal and a reportee in another portal. The newly developed chatbot smartly understands a user’s access level and delivers information based on their privileges, thus enabling information security.
How does it work?
- The user role is derived from the configuration of integrated portals at an organization.
- The bot enables configuration of custom roles to be assigned to different users.
- ‘Intents’ or in other words, the intention of your query, can be mapped to user roles, if any restrictions are necessary for the intent. For example, if you ask the bot to retrieve your personal details, the intent remains, ’fetch personal details’.
- The bot is designed to fit in any business function without any major change in its framework.
“Essentially, the same bot can be used in multiple business functions. How cool is that!”
What about conversation flows?
Every organization’s complex management structure demands diversified communication models and conversation flows. The bot developed at DS Innovation Lab specializes in processing massive amounts of information and delivering it in a user-friendly manner, with reaction times faster than humans.
What’s special about the bot’s conversations?
- It Identifies relevant information to deliver the message in plots, so that it’s fast and easy to consume for the user.
- It delivers information from data tables, which is far more efficient than normal tables as it allows filtering and sorting by default.
- It is designed to identify the exact query, regardless of dynamic nature of questions.
The workflow of this bot is currently tested for Human Resource and Sales functions. The testing has proved that it is flexible enough to be customized to fit multiple requirements of an organization.
What we learnt from testing the bot for HR and Sales business functions?
We took a typical scenario of an enterprise where payroll details of employees are managed on ERP while timesheets are managed on another portal.
- The User could access information from multiple portals seamlessly with ONE login ID
The bot was integrated with different portals of the organisation which enabled the user to access different portals from single login, based on the access configuration in the portal. The bot was integrated on Zoho ERP, Microsoft Dynamics 365, greytHR and Salesforce at the same time.
- New intents could be added easily without having to change the code or any re-deployments
Another highlight of our bot was the ease with which a new intent can be added. It was capable of handling the maintenance of existing intents as well as any new intents without having to change the code and hence avoiding re-deployment.
This was achieved easily through Admin Console using its user-friendly interface. The console allows adding new intents with a set of utterances (various ways of asking the same query is termed as utterance). Also, it allows configuring user roles that have access to relative intents.
- The bot could manage daily tasks like requests and approvals in a breeze
The bot can also be used for performing actions like booking a meeting in Outlook Calendar. All necessary information like meeting time, location and invitees will be taken from the user and the meeting will be scheduled. Also, managers can approve timesheets and leave requests from the bot directly. In short, just like we retrieve information from the portals it is possible to write back to the portals too.
- The bot displays Power BI Reports in chat UI
This is yet another striking feature of the bot. It is capable of bringing in Power BI dashboards into the bot UI within seconds of request.
- The chatbot has empathy
This chatbot is developed to handle conversations in a friendly manner. It keeps a log of all the unanswered questions and understands if any query or request can be attended to immediately.
The Bot engine is built with 3 different configurations.
- Python engine which uses sequence matcher for identifying the intents.
- Google dialogue flow which is Google’s powerful engine where we can configure the intents and slots.
- LUIS, which is Microsoft’s NLP engine that detects all the intents provided.
Chatbots are the friendliest way to execute tasks and deliver support to the users. The chatbot provides seamless assistance to simple enquiries and performs complex tasks effortlessly.
Employing and inculcating a chatbot in your business processes has proven to increase efficiency and improve the happiness quotient of employees. Feel free to reach out to use if you want to try this chatbot for your organization.
This post is written by Shreya Mohan from Data Semantics. She is an active developer who worked on building this chatbot. You can find her on LinkedIn or write to her via email@example.com for any further information.