Operational Transformation Through an Internal Customer-Response AI

• Operational visibility and requirements definition: We analyzed and structured the actual state of the client's inquiry-handling operations and the detailed configuration of the systems in use. Drawing on the current workflow and accumulated datasets, we designed the AI architecture and defined the requirements for each function.
• Driving the PoC (Proof of Concept): We drove system development based on the defined requirements. In coordination with a partner company, we established the structure for fine-tuning to improve the AI's response accuracy.
• Overall program management and execution: We led the consulting workstreams — including operational visibility and requirements definition — while coordinating with the partner company (Company A) responsible for AI architecture and demo build-out. We handled proposal development, framing of initial issues, and approach design, while leading overall program management (PMO).

[Challenges]

• Inquiries from dealers regarding products and maintenance occurred frequently, and staff were responding to each one individually by email — creating a high operational load.
• It was an urgent priority to deploy a system in which the AI automatically generated response drafts and staff could send them after a final review only — thereby reducing workload.

[Results]

• In partnership with a company holding AI expertise, we delivered the internal customer-response AI system project from proposal through delivery.
• We delivered AI architecture design and per-function requirements definition based on the workflow and datasets.