<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/">
  <channel>
    <title>DEV Community: Akshra</title>
    <description>The latest articles on DEV Community by Akshra (@axraa).</description>
    <link>https://dev.to/axraa</link>
    <image>
      <url>https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F4130399%2Fcc827af1-6327-4fea-8a00-9368202cddd1.png</url>
      <title>DEV Community: Akshra</title>
      <link>https://dev.to/axraa</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://dev.to/feed/axraa"/>
    <language>en</language>
    <item>
      <title>PhantomShell 👻</title>
      <dc:creator>Akshra</dc:creator>
      <pubDate>Sat, 31 Oct 2026 16:54:43 +0000</pubDate>
      <link>https://dev.to/axraa/phantomshell-gbg</link>
      <guid>https://dev.to/axraa/phantomshell-gbg</guid>
      <description>&lt;p&gt;This is a submission for the &lt;a href="https://dev.to/challenges/hacktoberfest-week4-2026-10-26"&gt;Hacktoberfest Open-Source AI Challenge Week 4: Ghost in the Machine&lt;/a&gt;.&lt;/p&gt;

&lt;h2&gt;
  
  
  👻 What I Built
&lt;/h2&gt;

&lt;p&gt;I built &lt;strong&gt;PhantomShell&lt;/strong&gt;, an AI-powered developer companion designed to make everyday development tasks easier.&lt;/p&gt;

&lt;p&gt;The inspiration behind PhantomShell was simple: developers frequently encounter confusing errors, unfamiliar terminal commands, and code-related problems. Instead of switching between multiple tools, I wanted to bring several developer utilities together in one place.&lt;/p&gt;

&lt;p&gt;I interpreted the &lt;em&gt;Ghost in the Machine&lt;/em&gt; theme by creating a ghost-themed developer assistant that works alongside developers and helps them understand what's happening behind the scenes.&lt;/p&gt;

&lt;p&gt;PhantomShell includes:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;CodeCrypt:&lt;/strong&gt; Code analysis to help developers understand their code.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Command Oracle:&lt;/strong&gt; Helps explain and analyze terminal commands.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Error Exorcist:&lt;/strong&gt; Assists in understanding errors and possible solutions.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;AI Chat Assistant:&lt;/strong&gt; An interactive assistant for development-related questions.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Session Journal:&lt;/strong&gt; Helps developers save notes and maintain development records.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Haunted Terminal:&lt;/strong&gt; A dedicated terminal-inspired interface.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;The goal was to build something more practical than a traditional chatbot: an AI companion focused on developer workflows.&lt;/p&gt;

&lt;h2&gt;
  
  
  🎥 Demo
&lt;/h2&gt;

&lt;p&gt;PhantomShell currently runs locally.&lt;/p&gt;

&lt;p&gt;You can explore the source code and follow the installation instructions in the README.&lt;/p&gt;

&lt;h2&gt;
  
  
  💻 Code
&lt;/h2&gt;

&lt;p&gt;The complete source code is available on GitHub:&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;GitHub Repository:&lt;/strong&gt; &lt;a href="https://github.com/akshrapsit1256/PhantomShell" rel="noopener noreferrer"&gt;https://github.com/akshrapsit1256/PhantomShell&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;README:&lt;/strong&gt; &lt;a href="https://github.com/akshrapsit1256/PhantomShell/blob/main/README.md" rel="noopener noreferrer"&gt;https://github.com/akshrapsit1256/PhantomShell/blob/main/README.md&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;The repository contains the frontend, backend, API routes, services, database configuration, and automated tests.&lt;/p&gt;

&lt;p&gt;Feel free to explore the code, suggest improvements, or contribute!&lt;/p&gt;

&lt;h2&gt;
  
  
  🛠️ How I Built It
&lt;/h2&gt;

&lt;p&gt;I built PhantomShell using a combination of modern web technologies and open-source AI tools.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;Frontend:&lt;/strong&gt;&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;React.js&lt;/li&gt;
&lt;li&gt;Vite&lt;/li&gt;
&lt;li&gt;JavaScript&lt;/li&gt;
&lt;li&gt;CSS&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;&lt;strong&gt;Backend:&lt;/strong&gt;&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;Python&lt;/li&gt;
&lt;li&gt;FastAPI&lt;/li&gt;
&lt;li&gt;Pydantic&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;&lt;strong&gt;AI Integration:&lt;/strong&gt;&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;Ollama&lt;/li&gt;
&lt;li&gt;Locally running open-source language models&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;&lt;strong&gt;Database:&lt;/strong&gt;&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;SQLite&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;&lt;strong&gt;Testing:&lt;/strong&gt;&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;Pytest&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;One of the important parts of this project was integrating Ollama to run AI models locally instead of depending entirely on external AI APIs.&lt;/p&gt;

&lt;p&gt;Running models locally gave me the opportunity to experiment with AI integration while keeping inference within my development environment.&lt;/p&gt;

&lt;p&gt;I organized the backend into separate routes, models, and services to keep the application modular and maintainable.&lt;/p&gt;

&lt;p&gt;I also implemented automated backend tests to verify important application functionality, input validation, and error-handling scenarios.&lt;/p&gt;

&lt;h2&gt;
  
  
  🌍 Why Does Open Innovation Matter?
&lt;/h2&gt;

&lt;p&gt;Open innovation matters because it makes AI more accessible to developers, students, and independent builders.&lt;/p&gt;

&lt;p&gt;For PhantomShell, using open-source AI tools and local inference made it possible to experiment with AI integration without depending entirely on paid, closed-source APIs.&lt;/p&gt;

&lt;p&gt;It also gave me more flexibility to experiment with models, customize prompts, understand the inference workflow, and build around my own requirements.&lt;/p&gt;

&lt;p&gt;I believe open-source AI encourages learning through experimentation. Instead of treating AI as a black box, developers get an opportunity to understand how the different components work together.&lt;/p&gt;

&lt;p&gt;That freedom was an important part of building PhantomShell.&lt;/p&gt;

&lt;h2&gt;
  
  
  👻 The Spirit of It
&lt;/h2&gt;

&lt;p&gt;The name PhantomShell represents the idea of having an intelligent presence alongside your development environment.&lt;/p&gt;

&lt;p&gt;While building software, developers constantly encounter unexpected errors, confusing commands, and problems that take time to understand.&lt;/p&gt;

&lt;p&gt;I wanted PhantomShell to feel like a little ghost living inside the developer's workspace — always ready to help investigate what went wrong.&lt;/p&gt;

&lt;p&gt;Building this project also helped me explore the practical side of AI integration, backend architecture, frontend development, and testing.&lt;/p&gt;

&lt;p&gt;For me, the most interesting part was turning an idea into a working application rather than just another AI chatbot.&lt;/p&gt;

&lt;h2&gt;
  
  
  🏆 Prize Categories
&lt;/h2&gt;

&lt;p&gt;Open-Source AI Challenge – Week 4: Ghost in the Machine.&lt;/p&gt;

</description>
      <category>devchallenge</category>
      <category>hf26challenge</category>
      <category>opensource</category>
      <category>hacktoberfest</category>
    </item>
    <item>
      <title>LazyFlow AI: Let AI Handle the Repetitive Work 🦥</title>
      <dc:creator>Akshra</dc:creator>
      <pubDate>Sun, 25 Oct 2026 15:36:02 +0000</pubDate>
      <link>https://dev.to/axraa/lazyflow-ai-let-ai-handle-the-repetitive-work-30oe</link>
      <guid>https://dev.to/axraa/lazyflow-ai-let-ai-handle-the-repetitive-work-30oe</guid>
      <description>&lt;p&gt;This is my submission for the &lt;strong&gt;&lt;a href="https://dev.to/challenges/hacktoberfest-week3-2026-10-19"&gt;Hacktoberfest Open-Source AI Challenge — Week 3: Slow and Steady&lt;/a&gt;&lt;/strong&gt;.&lt;/p&gt;

&lt;p&gt;What if the small, repetitive tasks that fill up our day could be handled by an AI assistant running on our own machines?&lt;/p&gt;

&lt;p&gt;That question inspired me to build &lt;strong&gt;LazyFlow AI&lt;/strong&gt;, a local AI-powered productivity assistant designed to reduce everyday digital busywork so we can spend more time on the things that actually matter.&lt;/p&gt;

&lt;h2&gt;
  
  
  💡 What I Built
&lt;/h2&gt;

&lt;p&gt;LazyFlow AI is a web application that uses a locally running AI model to help with everyday productivity tasks.&lt;/p&gt;

&lt;p&gt;Instead of manually going through long notes, organizing tasks, or drafting emails from scratch, users can get an AI-generated starting point and review it before taking action.&lt;/p&gt;

&lt;p&gt;Here are its main features:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;📝 Smart Summarizer:&lt;/strong&gt; Turns lengthy notes and text into concise summaries.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;✅ Task Generator:&lt;/strong&gt; Converts notes into actionable tasks while avoiding invented deadlines when the input doesn't provide them.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;✉️ Email Draft Assistant:&lt;/strong&gt; Helps draft emails without automatically sending them.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;📁 File Organizer:&lt;/strong&gt; Previews proposed file-organization changes and generates a PowerShell script, with optional direct execution.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;☀️ Daily Briefing:&lt;/strong&gt; Provides a convenient overview to help users get started with their day.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;⚙️ Model Settings:&lt;/strong&gt; Lets users manage their local AI model configuration.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;The goal isn't to replace human decision-making. It's to reduce the repetitive work that gets in the way of it.&lt;/p&gt;

&lt;p&gt;Whether you're a student managing assignments, a developer organizing work, or someone who simply wants a less cluttered workflow, LazyFlow AI aims to make everyday tasks a little easier.&lt;/p&gt;

&lt;h2&gt;
  
  
  🛠️ How I Built It
&lt;/h2&gt;

&lt;p&gt;I built LazyFlow AI using open-source technologies and local AI inference.&lt;/p&gt;

&lt;h3&gt;
  
  
  Technology Stack
&lt;/h3&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Frontend:&lt;/strong&gt; React 19, Vite, Tailwind CSS, and Lucide icons&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Backend:&lt;/strong&gt; Python, FastAPI, and Pydantic&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Database:&lt;/strong&gt; SQLite with SQLAlchemy&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;AI Model:&lt;/strong&gt; Google's Gemma 2 2B, running locally through Ollama&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Testing:&lt;/strong&gt; Pytest for backend tests&lt;/li&gt;
&lt;/ul&gt;

&lt;h3&gt;
  
  
  How It Works
&lt;/h3&gt;

&lt;p&gt;The React frontend provides the interface for interacting with the productivity tools.&lt;/p&gt;

&lt;p&gt;When a user requests a summary, task list, email draft, or another AI-assisted operation, the frontend communicates with the FastAPI backend.&lt;/p&gt;

&lt;p&gt;The backend processes the request and uses the locally running Gemma model through Ollama to generate a response. SQLite supports the application's data-storage requirements.&lt;/p&gt;

&lt;p&gt;This architecture separates the user interface, backend logic, data storage, and AI inference into distinct components.&lt;/p&gt;

&lt;p&gt;I tested the backend using Pytest, and all &lt;strong&gt;13 automated tests passed&lt;/strong&gt;. The tests cover the AI service, health endpoint, summarizer, and task-generation functionality.&lt;/p&gt;

&lt;h2&gt;
  
  
  🌱 Why Does Open Innovation Matter?
&lt;/h2&gt;

&lt;p&gt;For me, this is the most important part of LazyFlow AI.&lt;/p&gt;

&lt;p&gt;AI productivity tools can be useful, but they can also introduce concerns about recurring API costs, internet dependency, and sending personal information to external services.&lt;/p&gt;

&lt;p&gt;Using an open-weight model locally gives LazyFlow AI a different approach.&lt;/p&gt;

&lt;h3&gt;
  
  
  1. More Control Over Personal Data
&lt;/h3&gt;

&lt;p&gt;When local inference is used, text sent to the model can remain on the user's machine rather than being sent to a third-party hosted LLM API.&lt;/p&gt;

&lt;p&gt;This is particularly useful when working with personal notes, draft emails, and other private information. The overall privacy benefits still depend on how the application and its other features are configured.&lt;/p&gt;

&lt;h3&gt;
  
  
  2. No Paid Remote LLM API Required
&lt;/h3&gt;

&lt;p&gt;LazyFlow AI uses a locally running Gemma model instead of requiring a paid remote LLM API.&lt;/p&gt;

&lt;p&gt;Once the model and required dependencies have been downloaded, local inference can work without an internet connection, provided the necessary local services are running and the task doesn't depend on an external service.&lt;/p&gt;

&lt;h3&gt;
  
  
  3. Freedom to Experiment
&lt;/h3&gt;

&lt;p&gt;Open-weight models make it easier to explore different configurations and experiment with alternative models and workflows without being tied to one hosted AI provider.&lt;/p&gt;

&lt;h3&gt;
  
  
  4. Accessibility for Independent Developers
&lt;/h3&gt;

&lt;p&gt;A local AI workflow gives students and independent developers an opportunity to build useful AI-powered applications without making a paid LLM API a requirement.&lt;/p&gt;

&lt;p&gt;Open innovation isn't just about having access to model weights. It's also about having greater control over how the technology runs, how it is integrated, and how it can evolve.&lt;/p&gt;

&lt;p&gt;For LazyFlow AI, that freedom makes a locally run productivity assistant practical to experiment with and improve.&lt;/p&gt;

&lt;h2&gt;
  
  
  🦥 The Lazy Payoff
&lt;/h2&gt;

&lt;p&gt;LazyFlow AI is built around a simple idea:&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;Stop doing the same small tasks manually when a tool can help with them.&lt;/strong&gt;&lt;/p&gt;

&lt;p&gt;Instead of repeatedly rewriting notes into task lists, starting emails from a blank page, or manually preparing summaries, users can get an AI-generated starting point and review it.&lt;/p&gt;

&lt;p&gt;The assistant still leaves important decisions to the user. Email drafts aren't automatically sent, and file-organization actions can be previewed before execution.&lt;/p&gt;

&lt;p&gt;I haven't established a measured time-saving figure yet, so I won't claim a specific number of minutes saved. My next step is to test the application with real workflows and compare how long those tasks take with and without it.&lt;/p&gt;

&lt;h2&gt;
  
  
  🏆 Prize Category
&lt;/h2&gt;

&lt;p&gt;&lt;strong&gt;Best Use of Gemma&lt;/strong&gt; — integrating Google's open-weight Gemma model through local inference.&lt;/p&gt;

&lt;h2&gt;
  
  
  💻 Explore the Project
&lt;/h2&gt;

&lt;p&gt;&lt;strong&gt;GitHub Repository:&lt;/strong&gt; &lt;a href="https://github.com/akshrapsit1256/LazyFlow-AI" rel="noopener noreferrer"&gt;LazyFlow AI&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;The repository includes the frontend, backend, local AI integration, and tests.&lt;/p&gt;

&lt;p&gt;I'd love for other developers to explore the project, experiment with it, and suggest improvements.&lt;/p&gt;

&lt;h2&gt;
  
  
  🔮 What's Next?
&lt;/h2&gt;

&lt;p&gt;LazyFlow AI is still a project I want to improve.&lt;/p&gt;

&lt;p&gt;My next steps include testing it with more real-world workflows, improving its reliability, and exploring additional ways to make local AI genuinely useful for everyday productivity.&lt;/p&gt;

&lt;p&gt;Building this project gave me an opportunity to explore how open-weight AI can be integrated into a practical application rather than being used only for demonstrations.&lt;/p&gt;

&lt;p&gt;I wanted to build something that could be useful in everyday life while keeping the AI workflow under the user's control.&lt;/p&gt;

&lt;p&gt;Sometimes, being productive means knowing what &lt;em&gt;not&lt;/em&gt; to do manually. 🦥&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;What repetitive task would you automate first?&lt;/strong&gt;&lt;/p&gt;

&lt;h1&gt;
  
  
  ai #gemma #productivity
&lt;/h1&gt;

</description>
      <category>devchallenge</category>
      <category>hf26challenge</category>
      <category>opensource</category>
      <category>hacktoberfest</category>
    </item>
    <item>
      <title>CycleCare: a private, local cycle companion powered by Gemma and Ollama</title>
      <dc:creator>Akshra</dc:creator>
      <pubDate>Fri, 02 Oct 2026 18:53:23 +0000</pubDate>
      <link>https://dev.to/axraa/cyclecare-a-private-local-cycle-companion-powered-by-gemma-and-ollama-2j26</link>
      <guid>https://dev.to/axraa/cyclecare-a-private-local-cycle-companion-powered-by-gemma-and-ollama-2j26</guid>
      <description>&lt;p&gt;This is a submission for the Hacktoberfest Weekend Challenge: Build for a Friend&lt;/p&gt;

&lt;p&gt;What I Built&lt;/p&gt;

&lt;p&gt;CycleCare is a small web app that estimates where I am in my menstrual cycle, lets me log symptoms, mood and energy, and then asks a local AI model for general self-care and meal ideas for the day.&lt;/p&gt;

&lt;p&gt;The friend I built it for is me. I wanted a simple way to understand my cycle without creating an account or putting my health data on a server I don't control. I haven't handed it to anyone else, so everything in this post comes from my own use.&lt;/p&gt;

&lt;p&gt;It is a wellness companion, not a medical tool. It doesn't diagnose, recommend medication, or make medical predictions.&lt;/p&gt;

&lt;p&gt;Demo&lt;/p&gt;

&lt;p&gt;  &lt;iframe src="https://www.youtube.com/embed/RlaNDTgPgBc" width="710" height="399"&gt;
  &lt;/iframe&gt;
The video uses fake data only. It shows me setting a cycle date and food preferences, logging symptoms, checking the dashboard, and generating suggestions from the local gemma2:2b model.&lt;/p&gt;

&lt;p&gt;Code&lt;/p&gt;


&lt;div class="ltag-github-readme-tag"&gt;
  &lt;div class="readme-overview"&gt;
    &lt;h2&gt;
      &lt;img src="https://assets.dev.to/assets/github-logo-5a155e1f9a670af7944dd5e12375bc76ed542ea80224905ecaf878b9157cdefc.svg" alt="GitHub logo"&gt;
      &lt;a href="https://github.com/akshrapsit1256" rel="noopener noreferrer"&gt;
        akshrapsit1256
      &lt;/a&gt; / &lt;a href="https://github.com/akshrapsit1256/cyclecare" rel="noopener noreferrer"&gt;
        cyclecare
      &lt;/a&gt;
    &lt;/h2&gt;
    &lt;h3&gt;
      
    &lt;/h3&gt;
  &lt;/div&gt;
  &lt;div class="ltag-github-body"&gt;
    
&lt;div id="readme" class="md"&gt;&lt;div class="markdown-heading"&gt;
&lt;h1 class="heading-element"&gt;CycleCare — a local cycle companion&lt;/h1&gt;
&lt;/div&gt;
&lt;blockquote&gt;
&lt;p&gt;&lt;strong&gt;Built for the DEV / Hacktoberfest 2026 "Build for a Friend" challenge. The friend I built it for is me.&lt;/strong&gt;
&lt;em&gt;A small, local menstrual-cycle tracker with wellness and meal suggestions from an open-weight AI model running on my own laptop.&lt;/em&gt;&lt;/p&gt;


&lt;/blockquote&gt;
&lt;br&gt;


&lt;div class="markdown-heading"&gt;
&lt;h2 class="heading-element"&gt;🌸 About CycleCare&lt;/h2&gt;
&lt;/div&gt;

&lt;p&gt;&lt;strong&gt;CycleCare&lt;/strong&gt; is a small web app I built for myself to track my menstrual cycle, log daily symptoms, and get general self-care, hydration, and meal ideas from a local open-weight AI model (Gemma, run through Ollama). Everything runs on my own laptop, with no account and no cloud service.&lt;/p&gt;




&lt;div class="markdown-heading"&gt;
&lt;h2 class="heading-element"&gt;💡 The Problem&lt;/h2&gt;
&lt;/div&gt;

&lt;p&gt;I wanted a simple way to understand where I am in my cycle, but most cycle apps ask me to create an account and keep my health data on their servers. I didn't want that for something this personal.&lt;/p&gt;

&lt;p&gt;So I built CycleCare to run entirely on my own…&lt;/p&gt;&lt;/div&gt;


&lt;/div&gt;
&lt;br&gt;
  &lt;div class="gh-btn-container"&gt;&lt;a class="gh-btn" href="https://github.com/akshrapsit1256/cyclecare" rel="noopener noreferrer"&gt;View on GitHub&lt;/a&gt;&lt;/div&gt;
&lt;br&gt;
&lt;/div&gt;
&lt;br&gt;
&lt;br&gt;&lt;br&gt;
How I Built It&lt;br&gt;&lt;br&gt;
Frontend: React and Vite&lt;br&gt;&lt;br&gt;
Backend: Python and FastAPI&lt;br&gt;&lt;br&gt;
Database: SQLite, a single file on my machine&lt;br&gt;&lt;br&gt;
AI: Gemma (gemma2:2b), an open-weight model, run locally through Ollama

&lt;p&gt;The flow:&lt;/p&gt;

&lt;p&gt;I save my last period date, average cycle length and food preferences.&lt;br&gt;
I log a daily check-in with symptoms, mood, energy and notes.&lt;br&gt;
When I click "Generate Suggestions", the backend reads my data from SQLite and builds a structured prompt from my cycle day, phase, symptoms, preferences and restrictions.&lt;br&gt;
It sends the prompt to Ollama's /api/generate endpoint and asks for JSON back.&lt;br&gt;
The backend parses the JSON and checks its shape, and the UI shows self-care, hydration, movement, sleep, 2 to 3 meal ideas and one short piece of guidance.&lt;/p&gt;

&lt;p&gt;The system prompt tells the model not to diagnose and not to recommend medication. If Ollama isn't running, the app doesn't crash. It falls back to a generic built-in plan, labelled as coming from the built-in knowledge base, not the AI model.&lt;/p&gt;

&lt;p&gt;I used an AI coding agent to generate most of the code. I then ran the app myself and tested it through the full flow. The backend, the Ollama call and the UI all work together on my laptop.&lt;/p&gt;

&lt;p&gt;Limitations:&lt;/p&gt;

&lt;p&gt;It's built for one user and tested only by me.&lt;br&gt;
It needs a laptop running Ollama, so I can't deploy it for others to try.&lt;br&gt;
Cycle phases are simple estimates from the last period date and average cycle length. They aren't reliable for irregular cycles and shouldn't be used to plan or avoid pregnancy.&lt;br&gt;
A 2B model can be slow, generic or inconsistent, and the meal ideas are general suggestions, not nutritional advice.&lt;br&gt;
SQLite is unencrypted, so anyone with access to my laptop could read the data.&lt;br&gt;
Why Does Open Innovation Matter?&lt;br&gt;
My data stays on my machine. Cycle and symptom data is deeply personal. With local inference, my prompts never go to a third party.&lt;br&gt;
No account, no API key, no cost. After the one-time model download, it runs for free.&lt;br&gt;
I control the model. The model name is one config setting, so I can swap it for a different Gemma size or another open-weight model without changing the code.&lt;/p&gt;

&lt;p&gt;A closed API would have meant sending this data to a company's servers and paying per request.&lt;/p&gt;

&lt;p&gt;Safety&lt;/p&gt;

&lt;p&gt;CycleCare is a wellness companion and is not intended to diagnose, treat, or prevent medical conditions. AI-generated suggestions are general wellness information and should not replace professional medical advice.&lt;/p&gt;

</description>
      <category>devchallenge</category>
      <category>weekendchallenge</category>
      <category>hf26challenge</category>
    </item>
  </channel>
</rss>
