DEV Community: Ritu Yadav

The Lazy AI That Reads Your Empty Fridge So You Don't Have To

Ritu Yadav — Tue, 20 Oct 2026 13:29:47 +0000

“Channeling my inner sloth: why stand with the fridge door open wondering what to buy when an open-source vision model can do it for you?”

Welcome to Week 3 of the Hacktoberfest Open-Source AI Challenge! This week's theme is "Slow and Steady"—building something with open-source AI that rewards laziness and automates a tedious daily chore.

Meet my project: Fridge-to-Grocer, a local tool where I just snap a messy photo of my fridge or pantry, and an open-weight vision model analyzes what’s missing and automatically drafts my grocery shopping list
After a long day of staring at screens, the last thing you want to do is configure cloud apps or stream over heavy connections. The Wind-Down Companion makes relaxation completely effortless:

The Single Click: Run a simple local script right before bed.

The Local Generation: An open-weight model crafts a slow-paced, personalized narrative and ambient sound guide.

The Offline Peace: Runs 100% locally on your laptop with Wi-Fi turned off, letting you truly disconnect.

Core Stack:
Open-Weight Model: Running Gemma or Llama locally via Ollama for zero-latency text generation.

Local Python Runtime: Lightweight script for seamless offline execution.

💡 Why Open Innovation Matters Here
True Offline Freedom: No internet connection required. It runs safely on a laptop even when you're completely off the grid.

Complete Privacy & Calm: No telemetry, no cloud tracking, and no external servers listening in while you try to wind down.

Zero Cost: Set it up once and use it every single night forever without worrying about subscription fees or token limits.

Local AI YouTube Chapter Generator

Ritu Yadav — Sat, 17 Oct 2026 16:51:21 +0000

HeadingSync: Local AI YouTube Chapter Generator

What I Built

HeadingSync is a local-first utility for independent YouTube creators. It transcribes a video's audio and uses an open-weight model via Ollama to automatically generate clean, clickable YouTube chapter timestamps.

Why Open Innovation Matters

Data Privacy: Raw video and audio files never leave the creator's laptop, keeping unreleased content completely secure.
Zero Cost: Running open-weight models locally means $0.00 per video, avoiding expensive cloud transcription APIs.
Customization: Creators can tweak prompts or swap models to match their specific niche.

Landing My First User

I tested it with my friend Marcus, who runs a gaming YouTube channel. He dropped his 24-minute devlog audio into the tool, and it generated accurate chapters in under a minute. He loved it, said he'd gladly pay a $29 one-time fee for a GUI version, and asked for a direct YouTube API integration next!

AI Disclosure

AI tools were utilized during the drafting and coding process to assist with structuring the application logic and refining this write-up.

Building an Offline Wellness & Habit Coach with Open-Source AI to Help Devs Maintain Healthy Streaks

Ritu Yadav — Wed, 14 Oct 2026 18:13:43 +0000

Introduction

Remote workers and developers often get buried in screen time, leading to burnout and lost productivity. While there are plenty of wellness apps out there, users are often hesitant to share their daily sleep, screen habits, and personal routines with third-party cloud servers.

For Week 2 of the Hacktoberfest AI Challenge (Theme: Side Hustle), I built an offline, local-first wellness and habit coach powered entirely by open-source AI that keeps personal data completely private.

The Problem

Tracking daily habits or seeking productivity advice usually requires sending personal journaling data to centralized cloud APIs. For anyone conscious about digital privacy, sharing personal mental health or habit logs feels risky. We need a tool that lives locally on our machine and guides us without snooping.

How It Works (The Open-Source Core)

Local Inference: Powered by Ollama running an open-weight model locally on my laptop.
Open-Source Scripting: A lightweight Python script parses daily text logs and user streaks.
Offline Nudges: The agent analyzes the user's weekly habits and gently suggests offline breaks, hydration checks, and physical stretches without needing an active internet connection.

Why Open Innovation Matters Here

Absolute Privacy: Personal wellness data, sleep patterns, and daily struggles never leave the user's laptop.
Zero Subscription Cost: Users get an intelligent personal coach for free, avoiding monthly subscription fees charged by commercial SaaS wellness apps.
No Vendor Lock-in: Total control over the prompts and model parameters.

First User Feedback & Bonus

I tested this with a fellow community member who was trying to maintain a coding wellness streak:

Feedback: They loved the privacy aspect and the fact that it didn't ping any external servers.
Would they pay? They mentioned they'd love a polished desktop widget version for a small one-time fee to help them stay away from the screen.

Publishing & AI Disclosure

AI Disclosure: Portions of this project's code structure and documentation were brainstormed and refined with the help of AI coding assistants, built and tested locally for Hacktoberfest 2026.

What side hustle or wellness tool are you building this week? Let me know in the comments below! 👇

I Built an AI That Tells You to Stop Using AI and Go Outside 🌿

Ritu Yadav — Tue, 06 Oct 2026 15:30:43 +0000

I Built an AI That Tells You to Stop Using AI and Go Outside 🌿

What if the best AI experience was one that made you stop looking at your screen?

For Hacktoberfest 2026's Touch Grass challenge, I built NatureQuest AI — an outdoor activity companion powered by open-weight AI.

The idea is simple:

Use AI to spend less time with AI.

🌱 What is NatureQuest AI?

NatureQuest generates personalized outdoor challenges based on how much time you have, the activity you want, and your preferred difficulty.

For example:

"I have 45 minutes and want an easy outdoor activity."

NatureQuest can turn that into a real-world mission:

Walk for 20 minutes.

Find three different leaf shapes.

Stop and listen to your surroundings for two minutes.

Find something in nature you've never noticed before.

Write down three things you observed.

The goal is for the screen to be the shortest part of the experience.

🤖 Why Open-Source AI?

The core of NatureQuest uses an open-weight AI model instead of relying entirely on a closed AI API.

I used Gemma to generate personalized outdoor quests.

Using an open-weight model gives the project more flexibility for local inference, experimentation, privacy, and future customization.

📵 The No-Screen Mode

After generating a quest, users can start it and put their phone away.

Most apps try to keep you inside the app.

NatureQuest wants you to leave it.

🛠️ How It Works
User
↓
NatureQuest Web App
↓
Backend
↓
Open-weight Gemma model
↓
Personalized outdoor quest
↓
No-Screen Mode
↓
🌳 Go outside!

🌍 Why "Touch Grass"?

The obvious way to build an AI application is to make people spend more time interacting with it.

I wanted to reverse that idea.

The goal isn't to maximize screen time.

It's to help people get outside faster.

🔓 Why Open Innovation Matters

Using an open-weight model lets me experiment with local inference, different models, and future customization without being completely dependent on a closed AI API.

🚀 What's Next?

I'd like to add:

📸 Plant and nature photo identification

🐦 Bird-call identification

🗺️ Outdoor route recommendations

🌦️ Weather-aware quests

🌱 Gardening challenges

🏆 Outdoor achievements

But the core idea stays the same:

AI should sometimes help us use our screens less.

💭 Final Thoughts

We often measure AI products by how much they can do for us on a screen.

NatureQuest asks a different question:

What if AI's job was to get us away from the screen?

Build something with AI.

Then go touch some grass. 🌿

devchallenge #hf26challenge

AI Career & Skill Gap Analyzer

Ritu Yadav — Mon, 05 Oct 2026 15:29:10 +0000

AI career & skill gap anlyzer

overview

An open-source AI app built to analyzer student resumes,identify skill gaps,and recommend career roadmaps.

Features

*Automated resume parsing
*Real-time skill gap analysis
*Personalized learning suggestion

Teach Stack

*Python,Streamlit,Open-Source AI/NLP Models

My Quick AI Meal Planner: A Hacktoberfest Project for My Bestie

Ritu Yadav — Fri, 02 Oct 2026 17:05:06 +0000

This is a submission for the Hacktoberfest Weekend Challenge: Build for a Friend

What I Built

Demo

Code

How I Built It

Why Does Open Innovation Matter?

My Agent Session

Prize Categories

Helping My Best Friend Eat Healthy with AI: My Hacktoberfest Weekend Challenge

Ritu Yadav — Fri, 02 Oct 2026 17:02:39 +0000

Hey everyone! 👋 For this Hacktoberfest Weekend Challenge, I wanted to build something practical and personal for my best friend, who always struggles with finding healthy meal options that match her food preferences and busy college schedule.

Since I am exploring open-source AI and tools this year, I decided to build a simple, AI-powered meal planning helper that suggests quick, healthy recipes based on ingredients we already have at home.

What I Built:

A lightweight helper prompt/tool using open-weight models that:

Takes a list of leftover or available ingredients from the kitchen.
Considers dietary choices (like vegetarian or gluten-free preferences).
Generates a quick 15-minute recipe with steps.

Why Open Innovation Matters:

Using open-source tools and AI makes it so much easier for beginners like me to build small, customized utilities without worrying about expensive APIs or complex server setups. It runs smoothly and can be customized instantly!

My friend loved the idea of getting instant recipe suggestions, and it actually saved us time during our study sessions.

Happy hacking, everyone! 🎉

Building LinguaPulse: My Offline AI Language Tutor

Ritu Yadav — Fri, 02 Oct 2026 15:39:24 +0000

Building LinguaPulse: My Offline & Private AI Language Practice Partner 🌍

Hey everyone! For this Hacktoberfest Weekend Challenge ("Build for a Friend"), I built something for a friend who was struggling to practice a new language without feeling nervous or judged.

Meet LinguaPulse!

What is LinguaPulse?

LinguaPulse is a clean, local web app that acts as a patient AI language tutor. It chats with you in your target language (like Spanish, French, or German) and gently corrects your grammar mistakes in real-time, right inside the chat.

Why Open-Source AI Matters for This Project

Instead of using a heavy, paid cloud API (like OpenAI or Claude), LinguaPulse runs 100% locally on your computer using Ollama and open-weight models (like Llama 3 or Gemma).

Here is why this open-source approach made all the difference:

100% Data Privacy: My friend's chat history and personal practice data never leave their laptop and are never sent to any third-party server.
Zero Cost: Absolutely free to run, with no subscription fees or paid API tokens required.
Offline Friendly: It works completely offline, even without an internet connection.

How I Built It

I used Python and Streamlit to design a simple user-friendly interface, connected it via a local backend to Ollama, and set custom system prompts to turn the raw AI model into a supportive language coach.

It was an amazing weekend project to build, and my friend loves practicing on it without any performance lag or privacy worries!

Happy Hacktoberfest 2026! 🚀

Intro

Ritu Yadav — Thu, 01 Oct 2026 15:07:35 +0000

I am new to the community ,happy to be the part of dev.