Vanam by Mahaveer Varma

A developer built Vanam, an offline AI field guide that uses a device's camera, microphone, and GPS to identify birds and plants and generate field notes entirely on-device without uploading user data. The app runs on-device models for bird identification, plant classification, and field-note generation, passing results through a confidence policy before writing notes from a local database and saving them to a trail.

This is a submission for the Hacktoberfest Open-Source AI Challenge Week 1: Touch Grass https://dev.to/challenges/hacktoberfest-week1-2026-10-05 I built Vanam, an offline AI field guide for identifying birds and plants in the field. It uses the device camera, microphone, and GPS, and runs AI models directly on the device without uploading data. The results are checked through a confidence policy, written using facts from the local database, and saved to a trail. https://vanam-app-mahaveervarma.netlify.app/ https://vanam-app-mahaveervarma.netlify.app/ https://github.com/VegirajuMahaveerVarma/Vanam https://github.com/VegirajuMahaveerVarma/Vanam I built Vanam with separate layers for the core logic, data storage, and user interface. The app follows a pipeline of capture → identify → confidence check → generate field note → save to trail. I used on-device AI models for bird identification, plant classification, and field-note generation. Open innovation matters because it helps us create solutions by using different ideas and technologies. In Vanam, I combined AI, field data, and on-device models to create a useful offline nature guide while keeping the user’s data private. My agent session focused on building Vanam as an offline AI field guide, where the AI models work directly on the device. I used the session to connect identification, confidence checking, field-note generation, and saving discoveries into the trail. Gemma