Post-market surveillance in the medical device industry has long been a reactive process; manufacturers often respond to safety concerns only after adverse events or recalls occur. Basil Systems aims to change that. The company’s newly launched Safety Signalling platform harnesses artificial intelligence and large language models to proactively identify potential safety risks across millions of regulatory and adverse event records in real time. In this exclusive interview with Medtech Spectrum, Anthony Cirurgiao, Founder and CEO of Basil Systems, discusses how the platform is redefining safety monitoring, enhancing regulatory compliance, and helping medtech companies transition from reactive reporting to predictive safety intelligence.
Basil Systems’ Safety Signalling solution is described as a first-of-its-kind AI platform for proactive post-market surveillance. What inspired the development of this solution, and what gaps in the medtech industry does it aim to fill?
Medical device safety monitoring has historically been reactive—manufacturers often learn about potential product issues only after adverse events are reported or recalls are initiated. We saw a clear opportunity to change that.
Basil Systems developed Safety Signaling to bring associative intelligence into post-market surveillance. By analyzing millions of adverse event reports, recalls, and regulatory documents in real time, our AI identifies emerging patterns that human teams might not detect until much later. The goal is to give manufacturers, regulators, and investors an early-warning system—one that helps mitigate safety issues before they escalate.
This solution fills a critical gap in the medtech industry: the ability to continuously monitor and interpret global safety data at scale, turning post-market oversight from a compliance exercise into a source of strategic and competitive insights.
Traditional post-market surveillance is often reactive and resource-intensive. How does Basil’s AI-driven approach improve early risk detection and actionable insights for medical device manufacturers?
Traditional surveillance depends on manual review—analysts combing through event reports, inspection data, and recalls. The data is siloed, not indexed, and difficult to analyze. That process is slow, inconsistent, and prone to bias. Basil’s AI automates this process by continuously ingesting and analyzing data across multiple regulatory sources.
Our models identify anomalies and correlations in real time—for example, subtle increases in complaint types or patterns across similar device categories. This enables earlier detection of risk signals, often months before they would otherwise surface.
We also deliver these insights in structured, actionable formats that fit directly into manufacturers’ safety review workflows, enabling proactive intervention and quality improvement.
The platform leverages large language models to interpret complex narratives from adverse event reports, recalls, and regulatory filings. How does this semantic and contextual analysis differ from traditional keyword-based searches?
Traditional systems rely on keyword matching, which can miss nuance in regulatory data. For instance, a report describing “battery overheating” might not match one that says “thermal damage,” even though both refer to the same risk.
Basil’s large language model enhancements enable the LLMs to understand context, meaning, and relationships across millions of documents. They interpret unstructured narratives the way a human reviewer would—recognizing synonyms, causal patterns, and severity implications without “hallucination”.
This semantic analysis allows Safety Signaling to cluster related events, identify emerging issues, and present a unified view of risk that goes far beyond keyword search.
How do Safety Signaling’s outputs integrate with existing workflows, quality management systems, and regulatory processes to support decision-making and operational efficiency?
Integration was a core design principle. Safety Signaling delivers its outputs through APIs, dashboards, and automated alerts that connect seamlessly with manufacturers’ quality management systems (QMS), post-market surveillance tools, and regulatory compliance workflows.
Teams can customize thresholds, monitor specific product families, and generate audit-ready data and analytics aligned with ISO 13485 and FDA expectations. The platform’s structured outputs also support documentation for vigilance reporting, CAPA investigations, and periodic safety update reports.
By embedding associative insights and analytics directly into existing systems, Safety Signaling enables faster, evidence-based decision-making and reduces both compliance risk and operational overhead.
Could you share any early examples or case studies where the platform successfully identified a risk signal before it became a major safety issue?
The new Safety Signaling solution is in early deployment with a number of global MedTech companies, and we will share case studies as they are developed in the future.
Looking ahead, how do you see AI and predictive analytics transforming post-market surveillance and patient safety across the medtech industry over the next 3–5 years?
We’re entering a new era where post-market surveillance will evolve from a reactive compliance process into a proactive, value-driving capability. Over the next few years, AI will make it possible to detect, diagnose, and even forecast safety risks before devices reach patients.
Regulators are also encouraging this shift toward data-driven vigilance and real-time monitoring. As systems like Safety Signaling become standard, we’ll see fewer widespread recalls, faster corrective actions, and stronger trust between manufacturers, regulators, and patients.
Ultimately, proactive safety intelligence will become as essential to medtech operations as design controls or clinical validation—driving both safer outcomes and smarter business decisions.
