<?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: Babuya Singh Chauhan</title>
    <description>The latest articles on DEV Community by Babuya Singh Chauhan (@babuya_singhchauhan_0e02).</description>
    <link>https://dev.to/babuya_singhchauhan_0e02</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%2F4174979%2Fbb4fff65-8d27-4dfa-94b5-ef6dcd188422.png</url>
      <title>DEV Community: Babuya Singh Chauhan</title>
      <link>https://dev.to/babuya_singhchauhan_0e02</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://dev.to/feed/babuya_singhchauhan_0e02"/>
    <language>en</language>
    <item>
      <title>The Last Light: Keeping Memories Alive with Local, Open-Weight AI 🕯️</title>
      <dc:creator>Babuya Singh Chauhan</dc:creator>
      <pubDate>Mon, 26 Oct 2026 21:20:59 +0000</pubDate>
      <link>https://dev.to/babuya_singhchauhan_0e02/the-last-light-keeping-memories-alive-with-local-open-weight-ai-2h93</link>
      <guid>https://dev.to/babuya_singhchauhan_0e02/the-last-light-keeping-memories-alive-with-local-open-weight-ai-2h93</guid>
      <description>&lt;p&gt;&lt;em&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;/em&gt;&lt;/p&gt;

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

&lt;p&gt;What if the most meaningful use of AI wasn't answering questions, writing code, or automating work—but helping us keep the stories we never want to forget?&lt;/p&gt;

&lt;p&gt;That question inspired &lt;strong&gt;The Last Light&lt;/strong&gt;, a cinematic, privacy-first AI memory keeper built around the idea that memories deserve a place to live, without having to leave the devices where they are stored.&lt;/p&gt;

&lt;p&gt;The Last Light brings together the atmosphere of a candlelit remembrance altar, the intimacy of a personal journal, and the creative possibilities of open-weight AI.&lt;/p&gt;

&lt;p&gt;Here is what it offers:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;The Digital Ofrenda:&lt;/strong&gt; An interactive remembrance altar featuring candlelight, marigolds, photographs, and personal dedications.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;The Memory Archive:&lt;/strong&gt; A personal space for preserving names, dates, family stories, favorite foods, songs, meaningful places, and photographs.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Let the Memory Speak:&lt;/strong&gt; A local AI writing studio that transforms a memory into a first-person remembrance, a heartfelt letter, an atmospheric story, a poem, or a family chronicle.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Private by design:&lt;/strong&gt; Memories and photographs are stored locally in the browser, with tools for exporting and restoring personal archives.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Offline-oriented AI:&lt;/strong&gt; Open-weight language models can run directly in the browser on compatible hardware, rather than requiring every generation request to go through a hosted AI service.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;The goal isn't to make a machine pretend it can bring someone back. It's to give people a thoughtful way to preserve the details, stories, and little things that make a memory meaningful.&lt;/p&gt;

&lt;h2&gt;
  
  
  Demo
&lt;/h2&gt;

&lt;p&gt;🕯️ &lt;strong&gt;Experience The Last Light:&lt;/strong&gt; &lt;a href="https://the-last-light.ai.studio/" rel="noopener noreferrer"&gt;https://the-last-light.ai.studio/&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;Explore the experience, discover the digital ofrenda, and try the memory-writing studio on a compatible device.&lt;/p&gt;

&lt;p&gt;The first model download requires an internet connection. Local generation depends on browser support, available device resources, and the model being downloaded successfully.&lt;/p&gt;

&lt;h2&gt;
  
  
  Code
&lt;/h2&gt;

&lt;p&gt;💻 &lt;strong&gt;GitHub repository:&lt;/strong&gt; &lt;a href="https://github.com/babuyasinghchauhan-eng/the_last_light" rel="noopener noreferrer"&gt;https://github.com/babuyasinghchauhan-eng/the_last_light&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;The project is built with React, TypeScript, Vite, and WebLLM, with browser-native storage and a design focused on privacy and data portability.&lt;/p&gt;

&lt;p&gt;The repository includes the source code, project documentation, and an MIT license for the application. The model weights have their own respective licenses.&lt;/p&gt;

&lt;p&gt;Contributions, improvements, and ideas are welcome.&lt;/p&gt;

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

&lt;p&gt;I approached this as a local-first application rather than a conventional chatbot wrapped in a beautiful interface.&lt;/p&gt;

&lt;p&gt;The frontend uses React, TypeScript, Vite, and Tailwind CSS. The visual direction combines cinematic lighting, dark editorial layouts, candlelight, and marigold-inspired details to create an experience that feels more like entering a quiet digital sanctuary than opening another AI dashboard.&lt;/p&gt;

&lt;p&gt;For the AI layer, I used &lt;strong&gt;&lt;a href="https://github.com/mlc-ai/web-llm" rel="noopener noreferrer"&gt;WebLLM&lt;/a&gt;&lt;/strong&gt;, which enables language-model inference in the browser using supported WebGPU execution paths.&lt;/p&gt;

&lt;p&gt;The application is designed to work with quantized open-weight models, including options from the Qwen2.5 and SmolLM2 families. Model selection and actual inference depend on the model configuration and the device's capabilities.&lt;/p&gt;

&lt;p&gt;Instead of sending personal memories to a remote LLM endpoint, the writing studio passes the user's chosen memory and writing instructions to the locally running model. The model then generates a literary interpretation of those details.&lt;/p&gt;

&lt;p&gt;The prompts are designed to preserve the user's supplied information, avoid inventing biographical facts as truth, and treat first-person writing as a creative device rather than a claim of communication with a deceased person.&lt;/p&gt;

&lt;p&gt;For persistence, the application uses IndexedDB to keep memories and photographs in the browser. Export and restore functionality gives users a way to retain control over their own collection.&lt;/p&gt;

&lt;p&gt;A few technical realities are worth acknowledging:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;The initial application and model downloads require connectivity.&lt;/li&gt;
&lt;li&gt;Subsequent offline use depends on assets and model files being cached and available, along with a compatible browser and sufficient device resources.&lt;/li&gt;
&lt;li&gt;Local inference is not equally fast on every device, and WebGPU support can vary.&lt;/li&gt;
&lt;li&gt;Open-weight models have their own license terms, which must be respected independently of the application's MIT license.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;These constraints are part of the engineering challenge. Running AI locally is not magic; it is a trade-off between control, privacy, hardware requirements, and convenience.&lt;/p&gt;

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

&lt;p&gt;&lt;strong&gt;Because a memory is too personal to treat as just another API request.&lt;/strong&gt;&lt;/p&gt;

&lt;p&gt;Many AI experiences depend on sending prompts and data to a service controlled by someone else. That can be useful, but it also introduces external dependencies, usage limits, potential costs, and questions about where personal information goes.&lt;/p&gt;

&lt;p&gt;The Last Light explores another approach.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;1. Privacy should be an architectural decision.&lt;/strong&gt;&lt;/p&gt;

&lt;p&gt;Family photographs, personal stories, and intimate memories deserve care. By keeping memory storage in the browser and using local inference, the application can reduce the need to transmit this personal information to a third-party AI provider.&lt;/p&gt;

&lt;p&gt;The intention is simple: your stories should remain yours.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;2. The experience should not depend on a single AI provider.&lt;/strong&gt;&lt;/p&gt;

&lt;p&gt;With an open-weight model and a transparent inference layer, developers can inspect the architecture, adapt the prompts, and explore compatible models. The project is not conceptually tied to a single hosted model or API.&lt;/p&gt;

&lt;p&gt;That flexibility matters for a tool whose purpose is to preserve stories over time.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;3. AI should remain useful beyond a network connection.&lt;/strong&gt;&lt;/p&gt;

&lt;p&gt;After the required resources have been downloaded and cached, a compatible local setup can support offline generation. No recurring hosted inference request is inherently necessary for each story.&lt;/p&gt;

&lt;p&gt;That makes the experience more independent of network availability and external service uptime, although it still depends on the user's hardware, browser, and locally available model files.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;4. Open innovation makes experimentation accessible.&lt;/strong&gt;&lt;/p&gt;

&lt;p&gt;I can experiment with different model sizes, adjust the writing instructions, and make the experience behave differently without redesigning the whole product around a closed API.&lt;/p&gt;

&lt;p&gt;There are real trade-offs: local models may be slower or less capable than larger hosted models, and the first download can be substantial. But for this particular application, privacy, control, and the possibility of continued use are valuable design goals.&lt;/p&gt;

&lt;p&gt;Open innovation is not simply about making software free. It is about giving people more control over the technology they depend on.&lt;/p&gt;

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

&lt;p&gt;A favorite recipe. A familiar expression. A song that immediately brings someone to mind. A story told so many times that the family knows every word.&lt;/p&gt;

&lt;p&gt;Sometimes, what we fear losing isn't a grand story. It's one of these little details.&lt;/p&gt;

&lt;p&gt;That is the spirit behind The Last Light.&lt;/p&gt;

&lt;p&gt;The digital ofrenda draws inspiration from the remembrance traditions of Día de los Muertos, which honors the continuing connections people have with loved ones who have died. The intention is to approach that tradition with respect rather than reduce it to a Halloween effect.&lt;/p&gt;

&lt;p&gt;The AI writing studio adds a different dimension: it can help someone turn a scattered memory into something they can reread, share, edit, or preserve for the next generation.&lt;/p&gt;

&lt;p&gt;The result is still a human memory, shaped by the person who shared it. The AI is a writing companion, not a replacement for the person being remembered.&lt;/p&gt;

&lt;p&gt;And that distinction matters.&lt;/p&gt;

&lt;p&gt;Technology changes. Services disappear. Models evolve. But the stories people choose to preserve should not have to disappear with them.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;Some things disappear from the world. They don't have to disappear from our stories.&lt;/strong&gt;&lt;/p&gt;

&lt;p&gt;That is The Last Light: a small experiment in making AI feel a little more personal, a little more private, and a little more human.&lt;/p&gt;

&lt;p&gt;🕯️&lt;/p&gt;

</description>
      <category>devchallenge</category>
      <category>hf26challenge</category>
      <category>gemma</category>
      <category>ai</category>
    </item>
    <item>
      <title>Drift Studio: Less Posting, More Creating with Open-Weight AI</title>
      <dc:creator>Babuya Singh Chauhan</dc:creator>
      <pubDate>Wed, 21 Oct 2026 23:14:59 +0000</pubDate>
      <link>https://dev.to/babuya_singhchauhan_0e02/drift-studio-less-posting-more-creating-with-open-weight-ai-mpn</link>
      <guid>https://dev.to/babuya_singhchauhan_0e02/drift-studio-less-posting-more-creating-with-open-weight-ai-mpn</guid>
      <description>&lt;p&gt;&lt;em&gt;This is a submission for the &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;/em&gt;&lt;/p&gt;

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

&lt;p&gt;Meet &lt;strong&gt;Drift Studio — an open-AI-powered social content workspace built around one simple idea: less posting, more creating.&lt;/strong&gt;&lt;/p&gt;

&lt;p&gt;Coming up with ideas is often the easy part. Turning them into platform-specific posts, rewriting captions, organizing drafts, and planning a consistent content calendar is where the busywork begins.&lt;/p&gt;

&lt;p&gt;Drift Studio helps creators, indie hackers, developers, students, and small businesses turn one rough idea into a collection of social content drafts.&lt;/p&gt;

&lt;p&gt;Instead of starting from a blank page every time, users can build a campaign, generate content variations, edit their drafts, organize them on a visual content board, and plan their publishing dates in a calendar.&lt;/p&gt;

&lt;p&gt;Here's what it includes:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Campaign creation:&lt;/strong&gt; Turn a rough idea into a structured content plan.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Multi-platform content:&lt;/strong&gt; Create drafts tailored to platforms such as LinkedIn, Instagram, X, and other supported channels.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Content board:&lt;/strong&gt; Move drafts through Idea, Draft, In Review, and Ready to Publish.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Editorial calendar:&lt;/strong&gt; Organize content by date and plan the week ahead.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Draft editor:&lt;/strong&gt; Edit, duplicate, copy, and refine generated content.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Data ownership:&lt;/strong&gt; Export campaigns and drafts to JSON, CSV, or plain text.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;A distinctive interface:&lt;/strong&gt; An editorial design with bold typography, colorful artwork, and a responsive workspace.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;The goal isn't to automate creativity away. It's to remove enough repetitive work that creating becomes easier to sustain.&lt;/p&gt;

&lt;h2&gt;
  
  
  Demo
&lt;/h2&gt;

&lt;p&gt;Explore Drift Studio here:&lt;/p&gt;

&lt;p&gt;&lt;a href="https://drift-studio-two.vercel.app/" rel="noopener noreferrer"&gt;https://drift-studio-two.vercel.app/&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;The workspace includes a demo mode so visitors can explore its content-management workflow when local model inference isn't available.&lt;/p&gt;

&lt;h2&gt;
  
  
  Code
&lt;/h2&gt;

&lt;p&gt;GitHub repository:&lt;/p&gt;

&lt;p&gt;&lt;a href="https://github.com/babuyasinghchauhan-eng/Drift-Studio" rel="noopener noreferrer"&gt;https://github.com/babuyasinghchauhan-eng/Drift-Studio&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;The project is built with React, TypeScript, Vite, and Tailwind CSS.&lt;/p&gt;

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

&lt;p&gt;I built Drift Studio around a browser-based open-weight AI approach rather than making a proprietary, hosted generation API the foundation of the product.&lt;/p&gt;

&lt;p&gt;The inference architecture uses &lt;strong&gt;SmolLM2-135M-Instruct in an ONNX-based pipeline&lt;/strong&gt;, with Transformers.js for model execution and WebGPU acceleration where supported. A WebAssembly fallback and demo mode help accommodate environments where the preferred inference path is unavailable.&lt;/p&gt;

&lt;p&gt;The application separates its content workflow from model execution. This lets the workspace remain useful for editing, organizing, scheduling, and exporting drafts even when AI inference cannot run.&lt;/p&gt;

&lt;p&gt;The application also uses browser storage for local project persistence, avoiding the need to create an account or depend on a cloud database for its basic workflow.&lt;/p&gt;

&lt;p&gt;Model downloads and browser hardware support can affect the first run, so I don't claim that every device works completely offline immediately.&lt;/p&gt;

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

&lt;p&gt;For a tool that handles unfinished ideas, product announcements, and personal writing, where the content is processed matters.&lt;/p&gt;

&lt;p&gt;An open-weight approach creates opportunities that go beyond simply generating text:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;More control:&lt;/strong&gt; Model behavior, prompts, and the inference pipeline can be inspected and experimented with.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Privacy-conscious workflows:&lt;/strong&gt; When local inference is available, draft generation can happen in the browser without sending the content to a separate AI inference service.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Fewer API dependencies:&lt;/strong&gt; The core architecture doesn't require a proprietary text-generation API key for its local inference path.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;More flexibility:&lt;/strong&gt; The model and runtime can evolve independently of the content-management interface.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Accessible experimentation:&lt;/strong&gt; Developers can explore model behavior, deployment trade-offs, and different inference strategies without building everything around a closed endpoint.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;Open innovation isn't just about making AI cheaper. It's about giving developers more freedom to understand, adapt, and control the tools they build with.&lt;/p&gt;

&lt;p&gt;That distinction matters for Drift Studio: the AI should help people get work done without forcing the whole product to depend on a single proprietary service.&lt;/p&gt;

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

&lt;p&gt;The intended payoff is straightforward: spend less time rewriting the same idea for different platforms and more time actually building, learning, or creating.&lt;/p&gt;

&lt;p&gt;One campaign becomes an organized set of editable drafts rather than a pile of notes and unfinished tabs. The calendar and content board provide a place to manage the next step, while exports keep the user's work portable.&lt;/p&gt;

&lt;p&gt;It's a small attempt to make consistency easier without making the creative process feel like another full-time job.&lt;/p&gt;

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

&lt;p&gt;I'd like to keep improving model compatibility, the local inference experience, and the workflow for turning a single idea into useful content across multiple platforms.&lt;/p&gt;

&lt;p&gt;For now, Drift Studio is my contribution to the idea behind &lt;strong&gt;Slow and Steady&lt;/strong&gt;: build something that takes the repetitive work off people's hands while keeping them in control of the result.&lt;/p&gt;

&lt;p&gt;Thanks for checking it out!&lt;/p&gt;

</description>
      <category>devchallenge</category>
      <category>hf26challenge</category>
      <category>ai</category>
      <category>showdev</category>
    </item>
    <item>
      <title>Veriff</title>
      <dc:creator>Babuya Singh Chauhan</dc:creator>
      <pubDate>Sun, 11 Oct 2026 16:38:07 +0000</pubDate>
      <link>https://dev.to/babuya_singhchauhan_0e02/veriff-4k3a</link>
      <guid>https://dev.to/babuya_singhchauhan_0e02/veriff-4k3a</guid>
      <description>&lt;p&gt;*&lt;strong&gt;&lt;em&gt;Thinking what to do about this website, should I buy it?&lt;/em&gt;&lt;/strong&gt;*&lt;/p&gt;

</description>
    </item>
  </channel>
</rss>
