DEV Community: Tanvi Bajpai

LocalLaunch AI: A Local-First Marketing Assistant for Small Businesses Powered by Ollama | #Hacktoberfest

Tanvi Bajpai — Thu, 15 Oct 2026 09:52:40 +0000

This is a submission for the Hacktoberfest Open-Source AI Challenge Week 2: Side Hustle

What I Built

✳ LocalLaunch AI — Your Small-Business Marketing Assistant

LocalLaunch AI is a local-first AI marketing assistant designed for freelancers, home businesses, handmade-product sellers, and small online shop owners.

Small business owners often spend valuable time writing product descriptions, social media captions, and promotional messages. LocalLaunch AI helps simplify this process by turning a few product details into a complete, ready-to-edit marketing pack.

What it can generate:

Product listing titles and descriptions
Key product benefits
Social media captions
WhatsApp promotional messages
Relevant search keywords

Users can choose their brand voice, target audience, and preferred platform to create content that suits their business.

The goal is to help small business owners save time on repetitive marketing tasks so they can focus more on building their businesses.

Demo

Live Website: LocalLaunch AI

Code

GitHub Repository: LocalLaunch-AI

The project uses HTML, CSS, and JavaScript for the frontend, with a lightweight Node.js server connecting the interface to Ollama's local API.

How I Built It

I built LocalLaunch AI using:

HTML — Created the product input form and marketing dashboard.
CSS — Designed a clean, responsive interface with a calm, minimal visual style.
JavaScript — Handled form interactions, model selection, API requests, loading states, and generated results.
Node.js — Built a lightweight backend to communicate with the local AI service.
Ollama — Integrated an open-weight language model for local text generation.

The application uses Ollama's local API to generate marketing content. The default model is llama3.2, and users can select other compatible models installed on their machines.

Instead of sending prompts to a hosted AI provider, the app sends them to the configured Ollama service. With the default local configuration, inference runs on the user's own computer.

This project helped me explore how frontend development, backend integration, and open-weight AI can work together to solve a practical problem.

Why Does Open Innovation Matter?

Open innovation makes LocalLaunch AI more accessible and gives users greater control over their tools.

1. Privacy and control

Product notes are sent to the local Ollama service rather than a hosted AI API. This is useful for small businesses that prefer to keep their drafts and business information on their own machines.

2. Flexible model selection

Users can switch between compatible open-weight models available through Ollama. The project isn't tied to a single hosted model provider.

3. More predictable running costs

There are no per-generation charges from a hosted AI API. Users can generate content using their own hardware, although electricity, hardware, and setup still have costs.

4. Offline potential

After downloading a model, content generation can work without an internet connection, provided Ollama and the local application are running.

5. Freedom to experiment

Open models make it easier to experiment with different writing styles, prompts, and models for specific business needs.

For a small side hustle, avoiding dependence on a metered AI API can make repeated use more predictable. Open innovation makes this kind of practical AI tool easier to adapt and build upon.

My Agent Session

I used AI-assisted development to help design the project structure, integrate Ollama, build the interface, and document the setup.

I haven't attached a DevRelay agent session yet.

Prize Categories

Open-source AI / Local Inference

GreenBuddy: Turn AI-Powered Quests into Real-World Adventures

Tanvi Bajpai — Sun, 11 Oct 2026 15:00:36 +0000

Touch Grass with AI: Meet GreenBuddy — Less Scrolling, More Exploring!

This is a submission for the Hacktoberfest Open-Source AI Challenge Week 1: Touch Grass

What I Built

GreenBuddy — AI Outdoor Quest Generator

GreenBuddy is a browser-based AI project designed to encourage people to spend less time scrolling and more time exploring the world around them.

Sometimes, we want to step outside but don't know what to do, especially when we have limited time or motivation. GreenBuddy turns that problem into a fun little adventure by generating outdoor quests based on a user's available time, mood, and surroundings.

✨ Key Features
AI-powered outdoor quests: Generate activity suggestions tailored to your preferences.
Mood-based suggestions: Find activities that match how you feel.
Flexible time options: Discover ideas that fit into a short break or a longer outdoor adventure.
Nature journal: Record completed quests and keep track of your outdoor experiences.
Browser-based experience: Access the application through a website without installing a separate application.
Fallback suggestions: Keep the experience usable when AI generation is unavailable.

GreenBuddy is designed for students, busy people, and anyone who wants to build a habit of spending more time outdoors.

My goal is simple: make going outside feel less like a task and more like an adventure. 🌱

Demo

🌐 Live Website: Try GreenBuddy

💻 GitHub Repository: Explore the Source Code

Open the website, choose your preferences, and discover your next outdoor quest!

Code

🔗 GitHub Repository: https://github.com/bajpaitanvi57-alt/GreenBuddy

The project is built using:

HTML5 — webpage structure and user interface.
CSS3 — styling and responsive design.
JavaScript — user interactions, quest generation, and journal functionality.
Transformers.js — browser-based AI inference.
Browser localStorage — local storage for completed quest entries.

The source code is organized into three main application files:

index.html — the main interface.
style.css — the visual design.
app.js — application logic and AI integration.

The repository also includes a README with setup instructions and project documentation.

How I Built It

I built GreenBuddy using HTML, CSS, and JavaScript, with the aim of combining accessible web development with open-source AI.

For AI inference, the starter implementation uses Transformers.js and the Xenova/LaMini-Flan-T5-248M model. The application uses the user's selected preferences to create a prompt and attempts to generate a relevant outdoor quest in the browser.

Here's how the experience works:

The user selects their available time, mood, and surroundings.
JavaScript combines these preferences into a prompt.
Transformers.js loads and runs the configured model when supported.
The generated quest is displayed on the webpage.
The user can complete the activity and save it in the nature journal.

I also included fallback quest suggestions so that the application can still demonstrate its core experience if AI generation fails.

One of the interesting parts of this project was bringing AI into a simple, interactive website rather than building just another chatbot. The focus is on helping users take a real-world action, not simply generating more content for them to consume.

The model and its dependencies may need to be downloaded initially, so the browser-based setup should not be assumed to work completely offline without additional testing.

Why Does Open Innovation Matter?

Open innovation made it possible for me to experiment with AI as a developer without building an entire machine-learning system from scratch.

Using open-weight models and open-source tools gives developers the opportunity to explore how AI works, customise the experience, and build applications around specific problems.

For GreenBuddy, this matters in three ways:

Accessibility for developers

Tools such as Transformers.js make it easier to experiment with machine-learning models using familiar web technologies. This lowers the barrier to building an AI-powered project.

More control and flexibility

Working with a model that can run in the browser creates an alternative to sending every generation request to a hosted inference API. It also gives developers more control over how the AI experience is integrated into their applications.

Building for meaningful impact

Open-source tools allow developers to focus on creative applications of AI. Instead of using AI only to answer questions or automate digital tasks, GreenBuddy explores how it can encourage people to engage with the physical world.

For me, open innovation is about making experimentation more accessible and giving more people the opportunity to turn their ideas into working projects.

My Agent Session

I don't have a shareable DevRelay agent-session link to include at the moment.

This section can be updated if I publish a session demonstrating how GreenBuddy was built.

Prize Categories

Potentially relevant category:

Open-Source AI / Open-Weight Models: GreenBuddy uses an open-weight model through Transformers.js in its starter implementation.

Note: Confirm the official partner prize categories and eligibility requirements on the challenge page before submitting. Remove or adjust this entry if the category does not match the official rules.

🌿 GreenBuddy — Less scrolling. More exploring. One small quest at a time.

Thanks for checking out my project! I'd love to hear suggestions for making outdoor adventures more fun, accessible, and personalised.