Threat Tracker is an iOS app built with React Native that provides real-time threat intelligence by scraping the web for local news and analyzing data using Hyperbolic Labs' large language model (LLM). The app generates a danger level heat map based on geolocated threat data and allows users to manually input threat information by selecting a location and danger level. Users are notified when danger levels in their vicinity rise.

• Real-time Threat Scraping: Automatically scrapes the web for news articles related to potential threats in your area.
• AI-driven Analysis: Uses Hyperbolic Labs' LLM to extract coordinates and assess the danger level of incidents.
• Danger Level Heat Map: Displays a heat map that visually represents the current threat level in different areas.
• User Input for Threats: Users can contribute by manually adding threats, specifying the location and assigning a danger level.
• Push Notifications: Alerts users when there is a significant increase in danger levels around them.

• Frontend: React Native
• Backend: MongoDB hosted on Reflex.dev
• AI Integration: Hyperbolic Labs LLM for location and danger level processing
• Hosting: Reflex.dev

• Using FastAPI framework and adapted in Reflex
• Two endpoints connected to the MongoDB

#get request
import requests

url = "https://rxh-prod-api-threattracker.fly.dev/locations"

response = requests.request("GET", url)

#post information
import requests

url = "https://rxh-prod-api-threattracker.fly.dev/locations/"

payload = {
    "latitude": 1,
    "longitude": 1,
    "weight": 3,
    "summary": "summary goes here"
}
response = requests.request("POST", url, json=payload)

• LLM usage: Extracted coordinates using Hyperbolic Labs' LLM from News articles scrabed from web. Also the danger level of incidents were assessed using LLM from the news.