platoseedplatoseed
Real-time defensible oversight for clinical trials
We turn unstructured monitoring data - monitoring visit reports, EDC data - into structured deviation, safety, and quality findings for human review. Panoptive generates an audit-ready oversight trail as a natural byproduct. The platform aggregates findings and identifies systemic patterns across sites and subjects before they compound, and tracks every finding through decision, follow-up, and resolution.
Panoptive provides a decision infrastructure to enable real-time, defensible oversight of clinical trials. It converts monitoring reports and EDC data into structured findings, decisions, and an audit-ready oversight record to streamline inspection readiness.
Panoptive ingests monitoring reports and EDC data, surfaces findings, structures sponsor decisions with evidence and rationale, and generates a continuous audit-ready oversight record with hyperlinked citations to protocols, ICH-GCP clauses, and SOPs. It creates an inspection-ready package from real study data, links findings to actions, and enables cross-trial visibility and continuous reconstruction of the decision trail to defend oversight during inspections.
Who it’s for: Clinical operations leaders, quality/compliance heads, and sponsors overseeing multi-site trials in pharma, medtech, and biotech.
Pilot programs and live demos; mentions of pilot Run a pilot and SOC 2/11 compliance indicate market traction and progressing adoption.
Building the oversight engine for clinical trials @Panoptive. CS @ CMU, previously a tech lead on the Foundation Model Evaluation team at Apple.
Simulate, trace, and evaluate agent behavior in real-world environments
AgentHub launches a platform that simulates realistic environments to evaluate AI agents end-to-end, with multi-domain sandboxes (e-commerce, CRM, filesystem, browser, dashboards, etc.), full tracing in OpenTelemetry, and plug-and-play evaluation (LLM, human-in-the-loop, and rule-based). It targets teams building tool-use, browser, conversational, and computer-use agents to identify failures and generate fixes before deployment.
Formerly “AgentHub”
AI-Native FDA Regulatory Services
Automated clinical trial oversight and execution
