Spike Detector AI

Company: Spike Detector AI

Self-reported basis: The description is entirely self-reported by the author, unverified, and submitted as part of an OpenAI 2026 hackathon project on Devpost. No external corroboration or independent evidence exists for any claims made.

What it appears to be: A research prototype aimed at assisting neurophysiology professionals in reviewing EEG samples using AI, with a focus on educational support and preliminary screening. It is described as an undergraduate thesis project expanded into a platform during OpenAI Build Week.

What changed: The author states the original prototype was extended into a full research platform during Build Week, including improved GPT workflows, a professional website, and clearer documentation.

Single most important open question: Is there evidence of clinical validation or real-world use beyond the described undergraduate research context?

• The description states that Spike Detector AI is an educational research prototype.
• It consists of:
  • An undergraduate research thesis
  • A custom GPT for structured EEG review
  • A research website documenting methodology, validation, and future vision
• The system organizes technical observations about waveform morphology, epileptiform activity, and artefacts into a readable report intended for expert review.
• It is described as a human-in-the-loop system, not intended to replace neurophysiology professionals.

Confidence: Low — the product is described as a prototype, not a commercial offering.

• The author states that the project was inspired by a desire to explore how AI could support EEG interpretation without replacing experts.
• It positions itself as an educational tool and preliminary screening aid, not a diagnostic or clinical decision-support system.
• The project is framed as a research effort, not a product for sale.
• It emphasizes:
  • Structured educational support
  • Keeping expert judgment central
  • Responsible AI use in healthcare

Inference: The positioning has evolved from an academic thesis to a research platform, but no commercial or market-ready evolution is evident.

• The description states that the system is intended for neurophysiology professionals.
• It is designed to assist with EEG sample review, particularly for identifying potential epileptiform discharges.
• The target audience includes:
  • Clinical neurophysiologists
  • Students or trainees in neurophysiology
  • Researchers working with EEG data

Confidence: Low — no evidence of actual customers, user feedback, or market engagement beyond the author’s own research.

• Not evidenced.
• The description does not mention any pricing model, monetization strategy, or commercial use case.
• It is described as a research prototype, not a product for sale.

Inference: No business model is evident from the provided information.

• Built using:
  • Custom GPT
  • OpenAI models (specifically mentioned: GPT-5.6)
  • Human-in-the-loop workflow
• The system uses a heterogeneous EEG dataset including normal EEGs, epileptiform recordings, and artefact-containing samples.
• It was evaluated on:
  • 175 EEG trace images
  • 40 previously unseen EEG recordings
• Performance metrics reported:
  • 82.5% overall agreement
  • 70% sensitivity
  • 87% specificity

Inference: Technical implementation is described as research-grade, not production-ready.

• Not evidenced.
• No evidence of revenue, customers, or adoption beyond the author’s own thesis and prototype.
• The project is explicitly described as a research prototype, not a product in use.
• No data on user engagement, feedback, or iteration history beyond the initial development.

Inference: No traction or maturity signals are evident.

• Not evidenced.
• No mention of existing tools or platforms in the EEG or neurophysiology space.
• No competitive analysis or positioning against other AI or clinical tools is provided.

Inference: No competitive context is described, and no evidence of market presence or competition is available.

• The system is described as a research prototype, not a validated clinical tool.
• It is explicitly stated that the results are not intended as clinical validation.
• Risk of overpromising AI capabilities in a sensitive domain (healthcare) without real-world testing or regulatory alignment.
• The project is single-person, with no team or organizational structure described.
• No evidence of ongoing development, funding, or commercialization plans.

Inference: High risk of misalignment between stated capabilities and real-world utility, especially in clinical settings.

1. What is the extent of clinical validation or expert review that has occurred beyond the research thesis?
2. Has the system been tested with actual neurophysiology professionals in a real-world setting?
3. Are there any plans to move beyond the prototype stage, and if so, what are they?
4. How is the human-in-the-loop component implemented and maintained?
5. What are the legal or regulatory considerations for using AI in clinical EEG interpretation?