Skip to content

Repository files navigation

📱 Interface Screenshots

eDNA Sequence Upload River Health Dashboard Real-Time Alert Broadcasts
Screenshot (524) Screenshot (523) Screenshot (526)

RiverSense AI

AI-powered eDNA analysis for Indian river health monitoring

Team Creatus · Biothon 2026 · Marwadi University · Top 50 Nationally

RiverSense AI reads environmental DNA (eDNA) extracted from river water samples, classifies biological communities using AI, and generates a real-time Biodiversity Health Score — automatically alerting government agencies when a river is at risk.


The problem

River health monitoring in India is manual, infrequent, and slow. By the time lab results confirm a pollution event, the damage is done. No automated early-warning system exists for Indian agencies.


What RiverSense AI does

  1. Sequences — pulls real metagenomic eDNA data from river samples (source: NCBI-SRA, Ganga river, run SRR8837334)
  2. Classifies — runs Kraken2 taxonomic classification on organisms present in the water sample
  3. Scores — computes a 0–100 Biodiversity Health Score from the abundance profile
  4. Alerts — sends an automated email directly to government agencies when the score crosses a risk threshold

Automated government alerts

When a river scores below the healthy threshold, RiverSense AI automatically dispatches an alert email (via Gmail SMTP) to:

  • CPCB — Central Pollution Control Board
  • Jal Shakti Ministry — national water resource authority
  • WCCB — Wildlife Crime Control Bureau
  • NFDB — National Fisheries Development Board

No manual trigger needed. The pipeline runs, scores the river, and sends the alert — end to end.


Tech stack

Layer Technology
ML / AI DNABERT (fine-tuned), Kraken2, Gemini 1.5 Flash
Backend FastAPI, SQLAlchemy (SQLite), Redis
Frontend React, Tailwind CSS
Alerts Gmail SMTP
Data NCBI-SRA (SRR8837334)

Team contributions

TEAM LEADER : THIVISHA ROLE : AI/ML Pipeline (DNABERT FINETUNING AND AUTOMATION )

TEAM MEMBER: KOYEL GHOSH ROLE:BACKEND(FASTAPI +MONGODB)

TEAM MEMBER 2 : SWATI GHARA ROLE: FRONTEND

Frontend UI/UX details: See FRONTEND_CONTRIBUTION.md


Running the project

Frontend (Next.js)

npm install
npm run dev

Open http://localhost:3000 to see the app.

Backend

See backend setup instructions in /app or contact the backend lead.


Data source

Real metagenomic sequencing data from the Ganga river:

  • Study: SRP190174
  • Run: SRR8837334
  • Source: NCBI Sequence Read Archive

Built for Biothon 2026 — Environment & Biodiversity track

About

From Raw DNA to Real Action. An AI-powered platform that translates raw river eDNA datasets into instant, plain-English biodiversity intelligence and automatically alerts government agencies of ecological risks in minutes instead of years.

Topics

Resources

Stars

1 star

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages