Triageo

• 

  Triageo Dashboard (web)

• 

  Agent Triageo (native on slack)

Inspiration

after talking to sponsors at the sponsor bay, like TD and other companies, we realized that cybersecurity is super important and often overwhelming for teams. we wanted to build a tool that automatically triages logs, flags the most critical incidents, and provides actionable guidance, all in a way that’s easy to use via Slack and a web dashboard.

What it does

Triageo is a real-time AI-powered security triage system. it: • monitors log files for suspicious activity, failed logins, injections, and server errors • categorizes alerts by severity (low → critical) • posts alerts to Slack instantly • updates a live web dashboard for quick visualization and tracking • suggests next actions using an AI model

How we built it

• Slack bot using Python & Slack Bolt to parse logs and send messages
• AI integration via Cohere API for triage suggestions and severity classification
• Web dashboard with React + Express backend bridge for live updates using SSE
• Run/demo script to start bot, frontend, and backend together seamlessly

Challenges we ran into

• making real-time updates work simultaneously on Slack and the web dashboard
• React dev server can’t accept POSTs directly, so we needed a backend SSE bridge
• coordinating multiple services (bot, backend, frontend) for demo

Accomplishments that we're proud of

• working end-to-end system: logs → AI triage → Slack + web dashboard
• live dashboard shows all severities, categories, and suggested actions
• Slack bot supports real-time “why/how to fix” explanations via AI
• built a plug-and-play demo that judges can run easily with one script

What we learned

• coordinating multiple services in real-time is tricky but doable
• integrating AI into a practical workflow adds huge value to automation

What's next for Triageo

•       add more advanced threat modeling & automated mitigation suggestions

Built With

• cohere
• javascript
• python
• react
• slack