[AOGU Corporation] Overcoming the Limits of Human Care: "Koko-chan," an AI-Powered Companion Doll Implementing Person-Centered Dementia Care

Born from Personal Struggle: My Grandmother's Alzheimers and the Limits of Family Care

Hello, everyone. My name is Yukiyuki Sakata, Representative Director and CEO of AOGU Corporation. It is a pleasure to be here today.

At AOGU, we are developing a next-generation care system for dementia. The inspiration for this business comes from a deeply personal experience. My grandmother suffered from Alzheimer's disease when she was in her 80s (she later passed away at 86). As her condition progressed, she would repeatedly say things like "Please just kill me" or "I want to die" on a daily basis.

It is said that even as memory fades for dementia patients, their emotions remain intact. However, as her family, we lacked the mental and emotional bandwidth to truly support her during those extreme emotional crises. Even for family members who have studied caregiving, the constant emotional strain and patience required often lead to frustration, deteriorating the relationship. Witnessing this tragic gap and wishing to solve it since my student days is the core drive behind AOGU.

The international standard for dementia care is "Person-Centered Care"—an approach that respects the individual's life history and values, viewing the world from their perspective. Understanding the meaning behind each behavior and maintaining a stable, reassuring relationship is crucial. However, relying solely on humans for this constant emotional acceptance is unsustainable. Families face emotional exhaustion, and professional caregivers are bound by severe time constraints, leaving little time to sit and listen to an individual's anxieties. That is why we decided to leverage technology to provide the "persistent acceptance and active listening" that humans struggle to maintain.

Digitizing Elder-Care Expertise: Bridging "Infinite Patience" and "Contextual Understanding"

Conventional smart tools and monitoring robots possessed the "infinite patience" to interact with patients all day long, but they lacked "contextual understanding." They did not know the patient's life story, what they took pride in, or what triggered their anxieties. Consequently, they could not offer meaningful assistance when a patient experienced Behavioral and Psychological Symptoms of Dementia (BPSD).

To bridge this gap, we utilized generative AI to create a system that combines "infinite patience" with "deep contextual understanding." This resulted in "Koko-chan," an AI-powered companion doll.

Koko-chan features three core advantages:

1. UX Optimized for Dementia Patients: Screen-based devices like tablets and smartphones are too complex for dementia patients. However, companion dolls (doll therapy) are internationally recognized as highly effective for emotional attachment. Koko-chan is designed to initiate interaction naturally. Even if patients forget that Koko-chan is a talking robot, simply touching or holding the doll triggers conversations. For example, during evening hours when patients are prone to "sundown syndrome," Koko-chan will proactively check in to soothe their anxiety.

2. Digitized Expertise of Geriatric Care Specialists: We have digitized the "mental models" of expert geriatric nurses. Experienced nurses assess the elderly's medical history, past careers, and recent anxieties to preemptively ask, "Are you feeling anxious about something?" or "Are you in pain?" to build trust. We have structured this advanced clinical dialogue process into our AI algorithms.

3. Accumulation of "Care Optimization Data": By recording the patient's background, their current state, the specific prompts used by the AI, and the resulting therapeutic outcome, we continuously refine the conversation algorithms to deliver highly personalized, reproducible care.

Koko-chan has already been deployed to 15 monitor users. We have seen significant improvements: consistent prompting for hydration, prevention of daytime drowsiness, and a dramatic increase in daily verbal expression for patients who previously spoke very little. Most notably, severe BPSD symptoms like wandering or verbal outbursts have been successfully de-escalated with a single, timely prompt from Koko-chan.

B2B2C Expansion into the 3 Trillion JPY Dementia and MCI Market

Our business model targets a formal device release next April or May (we are currently in the MVP validation phase). Unlike existing high-end robots, we aim to lower the entry barrier by setting the device cost at just under 20,000 JPY, with a monthly subscription fee of approximately 7,000 JPY.

For our distribution channels, we are focusing on a B2B2C model, partnering with care managers and visiting nurses rather than relying solely on direct-to-consumer (D2C) marketing. Introducing Koko-chan through trusted professionals ensures high adoption and retention rates.

Regarding market potential, there are approximately 10 million individuals in Japan affected by dementia or Mild Cognitive Impairment (MCI), representing a Total Addressable Market (TAM) of 3 trillion JPY. We are primarily targeting patients with moderate symptoms (1 to 1.5 million people) who benefit most from conversational intervention. By capturing a 10% share of this segment, we aim to secure a Serviceable Obtainable Market (SOM) of 30 billion JPY.

We are currently raising funds for our official launch. We look forward to connecting with potential business partners, VCs, CVCs, and angel investors who share our vision. Thank you very much.

Q&A and Feedback

Mr. Ogawa (Commentator): Thank you, Sakata-san. Your personal drive to solve the emotional toll of dementia care through technology is highly commendable.

How does your business model, pricing, and distribution channel strategy differ from existing animal-themed or therapeutic robots (such as Paro, the robotic seal) that have entered the market in the past?

Mr. Sakata: Thank you. Existing therapeutic robots are extremely expensive, costing around 400,000 to 500,000 JPY, which makes them inaccessible for average households. We have significantly lowered this barrier by pricing the hardware under 20,000 JPY and offering a monthly subscription of 7,000 JPY.

Furthermore, traditional robots focus primarily on mechanical movements or sounds to stimulate attachment. They are not capable of "context-based verbal communication" to mitigate cognitive decline or de-escalate BPSD. Koko-chan's unique technical advantage lies in the combination of generative AI and clinical geriatric nursing expertise, enabling dynamic, two-way conversations.

Regarding distribution, since elderly patients cannot purchase or configure the devices themselves, we partner with care managers and visiting nurses (B2B2C). Entering the care ecosystem through trusted professional recommendations allows us to maintain high user engagement and gather high-value care data.

Mr. Ogawa: I see. A low-cost device coupled with a monthly subscription for digitized nursing expertise is a highly logical setup. Partnering with care managers is also a smart way to overcome the initial trust barrier.

For your SOM of 30 billion JPY, what percentage of the moderate-symptom market does that represent?

Mr. Sakata: It represents approximately a 10% share of the 1 to 1.5 million moderate-symptom dementia patients who are most suited for our conversational intervention.

Mr. Ogawa: As a solution that relieves the burden on caregivers while preserving the dignity of patients, I have high expectations for Koko-chan. I wish you the best for your official launch.

Mr. Sakata: Thank you very much. We will do our best.