I vibe coded an AI security system for my aging parents, they're safer
In today’s fast-paced world, technology plays a crucial role in enhancing our lives, especially for vulnerable populations such as the elderly. Srdjan Stakic, a former film producer, has taken it upon himself to create a unique AI security system designed specifically for his aging parents. This article explores Stakic’s journey in developing this innovative technology and the profound impact it has had on his family’s safety.
The Inspiration Behind the AI Security System
Two years ago, Stakic faced a life-altering challenge when he was diagnosed with stage four cancer. During this difficult period, he found that artificial intelligence (AI) became an essential tool for navigating the complexities of his medical situation. As he began to recover, he noticed that his parents’ health was declining, prompting him to assist them with daily tasks such as cooking, cleaning, and managing medical appointments.
Language barriers added to the challenges, as English was not their first language. Communicating with healthcare providers became a daunting task. To bridge this gap, Stakic utilized AI to document conversations with doctors and translate important information into Serbian for his parents. However, he soon realized that he wanted to create a more comprehensive solution that could monitor their safety and well-being.
What is Vibe-Coding?
Stakic’s journey into the world of technology began with a concept he refers to as “vibe-coding.” This approach combines intuitive design with coding principles to create user-friendly software. Despite having no prior coding experience, Stakic was determined to build a system that could observe his parents’ activities and assess their safety.
He began outlining his vision using AI chatbots like Gemini and ChatGPT to analyze the ethical and technical aspects of his idea. Through this process, he gathered input from his family on how they wished to be treated in various scenarios, such as during a fall or medical emergency. This feedback was crucial in shaping a flexible and responsive system.
Building the AI Security System
Stakic transitioned to using the platform Lovable, which provided a live development environment. This platform allowed him to describe his ideas and see them come to life in real time. He integrated various components, including the frontend, backend, database, and authentication systems, which he had previously been unaware of.
To train the AI, Stakic uploaded hundreds of instructional videos for healthcare providers and caregivers. He established a validation pipeline and labeled datasets using real-world caregiving footage, adhering to clinical benchmarks like Stanford’s C-I-CARE framework. This framework emphasizes respectful and professional interactions between caregivers and patients.
Features of the AI Security System
One of the standout features of Stakic’s system is its ability to detect falls in real time. He rigorously tested the technology by simulating falls in his living room to ensure the system could accurately recognize these incidents and respond appropriately. After several months of development and collaboration with an IT company, the system was able to send notifications to family members or emergency medical services (EMS) in the event of a fall, along with a summary of the individual’s health records.
Additionally, the system analyzes interactions between caregivers and Stakic’s parents. It can assess the professionalism of caregivers based on audio and video data, ensuring that the care provided meets established standards. Stakic emphasized that he did not want to invade his family’s privacy; therefore, the system only reviews footage when a concern is flagged, providing a brief summary of the incident.
Launching Alvis: A Startup for AI-Assisted Care
What began as a personal project for his family quickly evolved into a startup called Alvis. Stakic realized that many others could benefit from this technology, leading him to launch a company aimed at making the system accessible to a broader audience. Alvis is designed to detect falls, recognize exceptional caregiver behavior, and generate advocacy letters when issues arise.
The startup is currently in private beta and is accepting applications for its pilot program, which is set to launch on April 13. Alvis will operate on a subscription model, similar to existing camera cloud storage services, with a premium tier that includes AI-assisted analytics.
Real-Life Applications of AI in Healthcare
Stakic’s commitment to leveraging AI extended beyond his initial project. Recently, when his mother was hospitalized, he utilized AI in several ways to enhance her care. First, he employed AI as a real-time medical interpreter, allowing him to understand lab results immediately. This capability was crucial for staying informed about his mother’s condition.
Additionally, AI served as a clinical advocate, identifying discrepancies in medical assessments and flagging critical changes in her health. Stakic also used AI to translate updates into patient-friendly language, ensuring that his parents fully understood the information being conveyed.
Throughout this experience, Alvis continued to operate in his mother’s hospital room, providing an extra layer of security and monitoring her well-being with the consent of her care team.
The Future of AI in Elderly Care
Stakic’s journey highlights the transformative potential of AI in enhancing the safety and quality of care for elderly individuals. By combining technology with a deep understanding of caregiving, he has created a system that not only monitors physical safety but also fosters dignity and respect in caregiver interactions.
As Alvis prepares to launch, it represents a significant step forward in the integration of AI into healthcare, offering families peace of mind and empowering caregivers to provide better support for their loved ones.
Note: The information presented in this article is based on the experiences of Srdjan Stakic and his development of an AI security system for elderly care. The technology is still in its early stages, and ongoing developments may further enhance its capabilities.
