The problem Nirodh: Drug Sentinel solves

Community drug-crime reporting is fundamentally fragmented. Law enforcement currently relies on delayed word-of-mouth or unstructured phone calls, leaving officers without verifiable evidence or live situational awareness before arriving on scene. Meanwhile, narcotics tip lines are frequently overwhelmed by duplicate, low-quality, or malicious reports.

Nirodh: Drug Sentinel bridges the gap between citizens and law enforcement through a unified, intelligence-driven pipeline:
For Citizens: A friction-free, zero-login dashboard allowing anyone to submit geo-tagged photo or video evidence with descriptions in under a minute.
For Law Enforcement: A secure command console that automatically triages incoming intelligence using a two-tier verification engine (SHA-256 cryptographic hashing to drop exact duplicates + Gemini AI plausibility checks to filter spam).
Predictive & Field Surveillance: District-level heatmaps transform individual tips into actionable pattern-of-life intelligence, while an ad-hoc mobile camera feature converts any authorized officer's phone into a live surveillance node with real-time facial detection.

Challenges we ran into

Filtering Spam without Human Bottlenecks: Allowing instant anonymous media uploads created a high risk of platform abuse. Processing every submission manually would paralyze enforcement units.
Solution: We built a hybrid verification pipeline combining deterministic SHA-256 image hashing for fast, low-cost duplicate detection with an asynchronous Gemini AI pass to evaluate image-text plausibility before routing tips to an officer’s queue.
Mobile WebRTC & HTTPS Enforcement: Converting officer smartphones into real-time CCTV nodes required browser getUserMedia camera permissions, which strictly enforce secure HTTPS protocols and fail on standard local IP testing.
Solution: We routed our local environment through Cloudflare Tunnels to generate valid HTTPS endpoints for real-device field testing.
Cross-Origin Next.js Build Failures: Tunneling external traffic into the local Next.js development server triggered cross-origin resource sharing (CORS) blocks, severing Hot Module Replacement (HMR) WebSockets and breaking JavaScript chunk loading.
Solution: We updated server configuration settings to explicitly handle custom domain headers and allowed origin hosts, restoring real-time development sync across mobile test devices.