From Always On to Confident Service: 《捷捷寶寶粥》 Uses AI Customer Service to Ease Support Workloads with an Around 70% Resolution Rate
From choosing flavors and checking nutrition facts to heating and storage instructions, parents often have many details they want clarified before preparing a meal for their baby. For 《捷捷寶寶粥》, these everyday questions are also a key part of how the brand serves parents. But as inquiries keep coming in, how can a lean customer service team maintain response quality without monitoring messages all day?
In this Raccoon AI interview, we spoke with Jessie, Assistant Marketing Manager at 《捷捷寶寶粥》, about the team’s experience implementing AI customer service. From handling common inquiries with brand knowledge to preparing campaign Q&A in advance, Raccoon AI has gradually become part of frontline service—reducing the pressure on agents to stay online at all times while helping the team define a clearer division of work between humans and AI.
Detailed questions that begin with a bowl of baby porridge show why clear customer service roles matter
At 《捷捷寶寶粥》, customer service is currently handled primarily by one team member. The team had previously used automated reply tools to reduce workload through predefined scripts. However, script setup required significant time, and customers’ questions and scenarios often went beyond expectations—from nutrition facts, ingredients, and calories to a baby’s current eating situation. Some mothers would even share daily life details during the conversation, creating scenarios that were difficult to anticipate in advance.
This experience led the team to look for a solution that could understand different contexts and handle more everyday inquiries. Jessie shared that the primary expectation for adopting AI was to reduce the response burden on frontline staff, so they would not have to monitor the platform 24/7.
Partnering with Raccoon AI to handle 70% of everyday inquiries, so agents no longer need to watch every new message
After implementing Raccoon AI customer service, frontline service gained a new division of work: common questions are first answered by AI based on the brand’s provided knowledge, while cases that require human support or customers requesting a handoff are routed to agents. Jessie described the most noticeable change: “We finally no longer need to sit in front of the computer every day.” Heating instructions, nutrition facts, and product inquiries are all topics parents frequently ask about. After knowledge training and ongoing optimization, AI can resolve 70% of common questions.
Jessie also shared examples: customers often ask what to do if baby porridge is not finished, or how to handle products that have slightly thawed. For these usage-based questions, the team now only needs to organize its existing customer service knowledge into the system and review real conversations to confirm whether AI answers customers’ questions effectively. This significantly reduces the pressure on agents to stay online at all times.
Updating FAQs in advance, combined with response optimization and team collaboration, helps AI customer service keep pace with campaign cycles
As usage experience accumulated, 《捷捷寶寶粥》 gradually developed its own approach to maintaining the knowledge base. Jessie shared that the key is preparing campaign information before each promotion begins. The team uses ChatGPT to organize campaign details into around 20 to 30 FAQ items, covering question scenarios they may not have originally considered, and then imports them into the Raccoon AI knowledge base. As a result, AI can answer questions based on the latest campaign information, making campaign responses more accurate while improving the efficiency of data preparation.
Raccoon AI handles everyday inquiries, allowing agents to focus on urgent conversations that require immediate intervention
Looking ahead, Jessie hopes to further apply question categorization and tagging to better understand different customer segments through parents’ inquiries, providing a reference for future communication and product recommendations. This is especially valuable for complaints: if customers can avoid repeating the same details, human agents can move into issue resolution faster. From answering questions and supporting human handoff to informing marketing decisions, these expectations outline the next stage of AI adoption for 《捷捷寶寶粥》.