Built for the field. Built to be inspected.

AquaLens helps municipalities, environmental consultants, lake managers, NGOs, and researchers triage freshwater monitoring work. It runs a deterministic remote-sensing core first (Sentinel-2 retrieval, six water-quality spectral indices, a weighted risk score), then a five-agent Gemini layer — Coordinator, Scout, Historian, Analyst, Reporter — that adds context and writes the brief and citizen-facing summary.

Why it exists

Freshwater bodies suffer from pollution, eutrophication, and harmful algal blooms — and many regions can’t afford continuous in-situ monitoring. Satellite indices fill part of the gap, but extracting actionable insight from spectral data requires domain expertise. AquaLens reduces that gap by handling the pipeline end to end and producing an honest, advisory output the field team can act on.

Hackathon context

AquaLens was built for the AI Agent Olympics hackathon at Milan AI Week 2026, in the Agentic Workflows and Multimodal Intelligence tracks. The repository is open-source under the MIT License.

Author

Talha Abid — github.com/talhaabidj1.