Sanjay Kukreja has close to 14 years of work experience in the IT industry and his expertise ranges from consulting organization to IT/ITES companies. Throughout his career he has excelled in various roles including that of a business analyst, partner management and running P&L operations.
eClerx Services was looking for a chatbot solution that could support its employees, improve productivity and track all tickets-related closures while adhering to compliance.
The organization arrived at a Ticketing Intelligent Assistant (TIA). TIA interacts with employees over Lync (Skype for Business), understands their problem and tries to resolve them. Typical L1 support for an enterprise includes – IT helpdesk, HR helpdesk, admin helpdesk. The efficiency of these functions is often affected by the support team spending too much time on low-value easy-to-resolve tickets on repetitive issues.
TIA is an ideal solution for simple IT and HR related queries. Where TIA is not able to provide a solution, it logs a service request on the ticketing system for the support team to address the issue. The key components include:
1. Classification engine – Dedicated to understanding the user’s statement and its intent. It relies on machine learning and natural language processing (NLP) components to provide the results.
2. Training data engine – In order to train the bot, vast amounts of historical data must be fed into the system. This data resides on-premise within eClerx’s network.
3. Stemming engine – To extract the word formation/pattern using language-dependent rules.
4. RESTful web service – This service communicates and transfers data to/from the chatbot engine. The loose coupling of the bot combined with the messaging platform allows for scalability and modular design.
TIA has been effectively deployed to handle these simple requests, freeing up support teams to focus on more complex issues, thus improving productivity and efficiency.