Chatbot Development for Education Sector

The objective was to enhance user interaction, engagement on the website, and lead generation for an education sector client. The challenge was to develop a functional chatbot in 2017 when the technology was still emerging and best practices were largely undefined.

• Began with Amazon Lex for initial prototyping — results were suboptimal
• Evaluated IBM Watson before pivoting to Google QnA for superior NLP
• Engineered custom middleware to cache Google QnA responses, significantly reducing costly API calls
• Iterative fine-tuning post-deployment based on real user interactions

• Used Jira for meticulous sprint planning and project tracking throughout
• Led the project from conceptualization through deployment and post-launch optimization
• Coordinated post-deployment training and chatbot fine-tuning sessions
• Leveraged Excel for data analysis and conversation flow metrics