HormoneRx

HormoneRx is a self-reported AI-powered tool designed to analyze real-time doctor-patient conversations for potential hormonal drug interactions, with the goal of alerting physicians to dangerous or concerning medication combinations. The project was submitted by a single founder, Oleksandr Tarasov, as part of the OpenAI 2026 hackathon. It is described as using AI and speech-to-text technology to process medical conversations and flag contradictions in prescriptions.

The core commercial due-diligence question is: What is the actual utility and viability of this tool in real-world clinical settings?

There is no evidence of revenue, customers, or adoption beyond the author’s own description. The project appears to be a proof-of-concept prototype, not yet validated in practice. It lacks any indication of integration with existing healthcare systems or regulatory approval.

The description states that HormoneRx:

• Analyzes conversation between a doctor and a woman patient.
• Tracks prescriptions given by the doctor.
• Fires an alarm when there are dangerous hormonal interactions between medications.
• Operates in real-time, with a reported delay of 2–3 seconds.

It is built using AI technologies, including GPT-based models (specifically referencing "gpt-whisperer"), and involves indexing databases, knowledge graphs, and agent-based processing.

Inference: The tool appears to be an experimental clinical decision support system focused on women’s hormonal health. It is not a consumer-facing app but rather a diagnostic or advisory tool for doctors during consultations.

The author positions HormoneRx as:

• A real-time doctor conversation checker.
• Focused specifically on women's health and hormonal interactions.
• Designed to improve patient safety by detecting dangerous drug combinations.

It is described as a clinical decision support system that uses AI to flag contradictions in prescriptions during live consultations.

Inference: The positioning suggests an intent to reduce adverse drug events (ADEs) in women’s healthcare, particularly around hormonal therapies. However, the claim of real-time safety monitoring is unproven without evidence of deployment or validation.

The description states:

• The tool targets doctors.
• Specifically, it focuses on women as the patient demographic.
• It is intended for use during doctor-patient conversations, likely in clinical settings.

There is no mention of:

• Healthcare institutions or systems.
• Patients directly using the product.
• Any specific ICP (Ideal Customer Profile) beyond “doctors treating women.”

Inference: The ICP appears to be physicians or healthcare providers who treat women and are concerned about hormonal drug safety. However, no evidence supports whether this is a niche or broader market need.

The description does not include:

• Any information on pricing.
• Revenue model.
• Customer acquisition strategy.
• Monetization approach.

Not evidenced: No indication of how the product would be sold or who would pay for it.

The project is described as:

• Built with AI, including GPT-based models ("gpt-whisperer").
• Uses speech-to-context processing.
• Implements indexing and knowledge graph construction.
• Processes conversations in real-time with a 2–3 second delay.
• Uses queue management for agents working continuously.

Inference: The technical stack suggests an early-stage prototype using AI tools for natural language understanding and clinical data analysis. It is not clear if this has been scaled or deployed beyond the hackathon context.

The description states:

• This was a hackathon submission.
• The team size is 1 person.
• No mention of:
  • Customers.
  • Revenue.
  • Product usage metrics.
  • Deployment in healthcare systems.
  • Regulatory or clinical validation.

Not evidenced: There is no evidence of traction, adoption, or product maturity beyond the initial concept and prototype phase.

The description does not include:

• Any mention of competitors.
• Existing tools in the clinical decision support space.
• Market size or competitive positioning.

Not evidenced: No information on how HormoneRx compares to other tools or systems addressing drug interaction warnings or women’s health.

Key risks and open questions based on the description:

• Single-founder project: Limited team capacity for development, scaling, or regulatory compliance.
• Unvalidated clinical use case: No evidence of real-world testing or integration with healthcare workflows.
• AI hallucination and accuracy concerns: The author explicitly mentions challenges with hallucination and validation.
• No revenue or customer data: The tool is not yet monetized or adopted.
• Hackathon prototype: Likely not production-ready or scalable.

Inference: The project is in a very early stage, and the clinical utility of such a system remains unproven. Regulatory, ethical, and technical risks are high without further development or validation.

1. What specific hormonal therapies or medications does the tool currently support?
2. How does it validate its findings? Is there any benchmarking or clinical testing?
3. Has it been tested in real-world doctor-patient interactions, or is it purely a prototype?
4. What are the regulatory and compliance considerations for deploying such a system in healthcare?
5. How does it handle privacy and data security, especially with sensitive health information?
6. What is the plan for expanding the database of drug interactions beyond the current scope?