CareOrbit

The description states that CareOrbit is a patient-facing mobile companion for people managing chronic conditions such as diabetes and hypertension. It provides four destinations:

• Today: Personalized medication, blood-pressure, and glucose reminders.
• Chat: Plain-English health questions and photo uploads for meter readings or meals.
• Health records: Simple views of the patient’s readings, medication confirmations, trends, and appointments.
• Doctor advice: A clear comparison between what the AI suggested and what the clinician ultimately approved.

The system uses GPT-5.6 through an OpenAI API gateway to handle conversational health questions, structured summaries, multimodal analysis (of synthetic blood-pressure and glucose meters), meal-image understanding, retrieval query rewriting, and validated structured outputs.

It implements a safety-gated retrieval workflow for general education questions using a source-verified knowledge base via OpenAI File Search. For safety-sensitive queries, it uses deterministic rules before calling GPT-5.6.

The product is built with FastAPI, SQLite, HTML, CSS, JavaScript, and Codex.

Evidence Self-reported by the author; no independent verification or data on actual use.

The author positions CareOrbit as an AI-powered daily care tool that reduces routine clinician workload while ensuring patient safety through human-in-the-loop design. The goal is not to replace medical judgment but to support patients with reminders, information access, and appropriate escalation paths.

Key claims include:

• Providing timely daily support for chronic care patients.
• Reducing the burden on clinicians by handling routine tasks.
• Ensuring AI suggestions are reviewed and approved by clinicians.
• Not allowing AI to independently change medication or override clinician decisions.

The positioning evolves from a general idea of "AI-powered health support" into a specific focus on patient-facing interfaces with structured clinician handoffs.

Evidence Self-reported; no external validation or market positioning data provided.

The target customer is described as people managing chronic conditions such as diabetes and hypertension. The product is intended for patients who need daily reminders, health record access, educational support, and communication with clinicians between appointments.

It is explicitly stated that the interface is designed for patients only, not clinicians, though clinician decisions are visible within the "Doctor advice" section.

Evidence Self-reported; no evidence of customer segmentation or user feedback beyond the author’s own experience.

There is no evidence in the description of a business model or pricing structure. The project is presented as a hackathon submission with no indication of monetization, subscription plans, or revenue streams.

Evidence Not evidenced.

The system is built using:

• FastAPI
• SQLite
• HTML, CSS, JavaScript for mobile interface
• GPT-5.6 via OpenAI API gateway
• Codex for development assistance

It includes:

• Structured SBAR handoff summaries
• Multimodal analysis of synthetic blood-pressure and glucose meters
• Meal-image understanding
• Retrieval query rewriting
• Validated structured outputs
• Safety-gated retrieval workflow with source verification
• Deterministic rules for risk classification
• Automated tests (23)
• Synthetic patient data (20 fixed-seed patients, 90 days of observations)

The author notes that the mobile interface required several refinements to balance clarity and usability.

Evidence Self-reported; no evidence of production deployment or scalability beyond prototype.

There is no evidence of traction, customers, or adoption. The project is described as a single-person hackathon submission with synthetic data and automated tests. No real-world usage, user feedback, or performance metrics are mentioned.

Evidence Not evidenced.

The description does not mention any competitors or competitive landscape. It focuses solely on the author’s own solution without reference to existing tools in the chronic care or AI health assistant space.

Evidence Not evidenced.

• Single-person development: The entire project was built by one person (Gloria J), raising questions about scalability, long-term maintenance, and team structure.
• No real-world testing: No evidence of usability testing with actual patients or clinicians.
• Unverified clinical safety: While the author claims safety boundaries, there is no evidence of regulatory compliance, privacy safeguards, or formal safety evaluation.
• Prototype-only status: The project is described as a hackathon submission and lacks production-grade features or deployment history.
• No commercial viability: No mention of monetization, partnerships, or business strategy beyond personal development.

Evidence Self-reported; no third-party validation or risk assessment data.

1. What specific clinical workflows or integrations are planned for future versions?
2. Has the author conducted any usability testing with actual patients or caregivers?
3. How does the team plan to ensure compliance with healthcare regulations (e.g., HIPAA, FDA)?
4. Are there plans to validate the educational content against authoritative sources?
5. What is the roadmap for moving from prototype to production-ready system?
6. How will patient authentication and data security be implemented at scale?

Inference These questions are based on the self-reported nature of the project and its lack of real-world validation or commercial strategy.