Growth in conversions
Increase in Customer Engagement
HealthifyMe makes day to day health management easy for its users. It offers personalized health guidance through their AI chatbot named Ria. It also provides personalized diet, workout suggestions as well as access to expert coach guidance.
Managing and improving health conditions requires a trusted partner who understands your health needs deeply. Gaining this trust in a short interaction over a chat is challenging. Hence, Healthify’s pre-sales agents were well-trained to generate customer trust in a short interaction.
However, as Healthify kept expanding into wider domains, including NRIs, the chat volume increased drastically. Due to increased workload, the level of personalization in replies decreased while user wait time increased. Additionally, timezone mismatch led to reduced agent availability for NRI customers.
All these challenges meant that the conversion rate remained stable from 12 to 14%, which was below Healthify’s targets.
Given enough context, AI agents are well suited to understand the customer needs and generate personalized messages. Also, due to their round the clock availability, timezone mismatch is never a challenge for AI agents.
For the current use case, a single AI agent seemed inadequate to hold fruitful user interactions. So we decided to setup a multi-agentic architecture. 3 inter-linked AI agents were setup each tasked with specific aims. Each new chat was passed through a cascade of agents:
The agents were provided with basic user details like name, age, BMI, medical conditions. This helped agents ask contextual questions, offer personalized plan recommendations and setup free consultation calls.
Additionally, AI agents were also capable of the following:
Round the clock AI agent availability along with personalization increased conversion rates to 18% (from 12-14% earlier).
