Team Idea
Using AssemblyAI's real-time speech-to-text, SentinelLine transcribes a spoken incident report as the user talks. An LLM then extracts the key structured fields — what happened, which systems or accounts are affected, and any urgency signals — while a lightweight retrieval layer matches the description against common attack patterns (phishing, credential stuffing, malware, etc.) to suggest a likely category and severity. The agent responds by voice with a confirmed, triaged ticket and immediate next steps, so the reporting user doesn't have to know security jargon to get useful guidance.
Submission
Team Leader hasn't made a submission yet