How Manufacturing Major Saved $6000
per Week through AI-Driven Chatbot

Client Challenges

• The company’s main challenge was online branding and they aimed to promote their brand through content marketing.
• To capture the details of a large number of distributors and retailers, the client was using an in-house enterprise management system
• The client also used it to keep an eye on manufactured PCBs
• However, the existing system was old and designed as per the then requirements
• It failed to incorporate the details of the growing network of distributors and retailers

This inefficiency resulted in –

• Distributors and retailers started to contact the client more frequently for enquiring about account balance, shipment status, payment dates
• The number of such calls went up to 500. In the past, it hovered around 200 – 300
• Some dedicated and crucial resources were involved in only addressing the support issues, which upped the cost by 15%
• Since client was spending more energy on resolving support issues rather than focusing on mission-critical tasks, productivity was reduced by 20-25%

Technology Stacks

• SAP
• Microsoft BOT Framework
• Microsoft SQL Server
• Azure Cloud

Our Solutions

• The main pain point of the client was the increasing number of calls, as they were consuming a significant time of their critical resources
• The client wanted to reduce the calls along with ensuring the increased information availability to distributors/retailers
• Automating the support service was the most viable and feasible solution
• We proposed and developed AIML based Chatbot solution (AMI)

The solution saw following implementations –

• Microsoft LUIS framework used to develop Chatbot solution absorbed user inputs and identified intents and entities to be used by Chatbot
• Azure hosted SQL Server with the UMLS Meta thesaurus queried the database using identified intents and entities
• The backend component processed responses and guided interactions with the user, as well as acted as the middleman between the NLP and Database