DEV Community: Sakshi Bhosale

Resurrecting Family Stories: Why I Built a Local-First Oral History Companion 📜🕯️

Sakshi Bhosale — Tue, 27 Oct 2026 20:22:44 +0000

This is a submission for the Hacktoberfest Open-Source AI Challenge Week 4: Ghost in the Machine

What I Built

The Legacy Ledger is a private, local-first web application designed to capture, structure, and preserve life stories, spoken memories, and family traditions.

I wanted to interpret the "Ghost in the Machine" theme by looking at how we preserve the literal "voices of our past"—the fading memories and anecdotes of elders that often get lost in ephemeral chat apps or locked away in proprietary cloud servers. Instead of building another spooky terminal wrapper, I wanted to build something that honors family history with warmth, permanence, and absolute privacy.

Demo

Live App: Check out the live deployment on Vercel
GitHub Repository: The Legacy Ledger on GitHub

Code

The application is built cleanly around a modular React component structure with local storage persistence, ensuring your family's personal reflections never leave your device.

sakshi-db / the-legacy-ledger

A private, local-first oral history companion to record, organize, and archive family life stories.

📜 The Legacy Ledger

Preserving family histories, spoken memories, and ancestral wisdom—privately, locally, and timelessly.
Built for Hacktoberfest 2026: Week 4 – Ghost in the Machine (Halloween & Día de los Muertos Theme) 🎃👻

🕯️ Overview

The Legacy Ledger is a local-first, privacy-focused oral history companion designed to capture, structure, and preserve life stories and family traditions.

In a digital age where personal reflections and family records often get lost in ephemeral chat apps or swallowed by third-party clouds, The Legacy Ledger brings back the warmth of an heirloom journal—giving your stories a structured, permanent home without compromising your data privacy.

✨ Key Features

🎙️ Spoken & Written Memory Capture: Seamlessly record unstructured audio notes or type out personal reflections and family anecdotes.
🏷️ Smart Structuring & Categorization: Automatically tag and organize memories into intuitive life eras (Childhood, Family Traditions, Milestones, Wisdom).
⌛ Visual Family…

View on GitHub

How I Built It

The app is built using React, Vite, Tailwind CSS, and Lucide React, styled deliberately with a warm archival aesthetic (parchment tones, deep slate typography, and muted amber accents) to feel like a timeless physical journal rather than a cold software dashboard.

While it's designed to seamlessly plug into local-first open-weight model architectures for auto-tagging and summarization, the frontend architecture focuses on zero-friction usability so family members of any technical background can easily record, browse, and structure their life stories.

Why Does Open Innovation Matter?

When it comes to family history, voice notes, and emotional memoirs, data privacy is non-negotiable.

Using a closed API or a third-party corporate cloud means sending deeply personal family memories, voices, and historical reflections onto external servers where they can be scraped or monetized. An open-source, local-first approach ensures that ancestral stories remain completely sovereign, secure, and private to the user. Open innovation makes true data ownership possible for intimate human stories.

The Spirit of It

Building The Legacy Ledger gave me a chance to step away from complex backend bloat and focus on the emotional resonance of software. There's something special about writing code that doesn't just solve a dry technical bug, but actually builds a digital home for the stories we can't afford to forget.

Crafted with ❤️ for Hacktoberfest 2026.

🦥 Solace: Productivity That Adapts to Your Energy

Sakshi Bhosale — Wed, 21 Oct 2026 19:55:57 +0000

This is a submission for the Hacktoberfest Open-Source AI Challenge Week 3: Slow and Steady

What I Built

As a college student juggling coursework, exams, lab submissions, personal projects, and daily habits, there are days when I wake up feeling completely overwhelmed. Facing a long list of 15 checkboxes on a low-energy day causes instant decision paralysis and guilt.

I built Solace Sanctuary—an adaptive, low-friction student sanctuary and life tracker designed specifically for low-battery days.
Instead of treating productivity like a rigid checklist, Solace works with your energy level:

🔋 Vibe Check (Energy Scaling): Switch between High Energy, Normal, and Zombie Mode. Selecting Zombie Mode automatically downscales daily targets (e.g., changing 3 DSA problems to 1 quick snippet review, or 1 hour of workout to 10 minutes of stretching).
🧘 One-Next-Step Focus Card: Shields you from feeling overwhelmed by showing ONLY ONE micro-task at a time. Pressing "I'm Too Tired / Break It Down" splits it into a tiny 2-minute step.
📝 Smart Mind Dump & Auto-Tracker: Write or paste messy, unfiltered thoughts, daily notes, or voice transcripts. The system parses the input and auto-sorts it into a daily journal entry, checks off daily tasks (assignments, DSA, workout, anything!), and queues tomorrow's tasks.
🎨 Custom Aesthetic Themes: Includes 3 switchable themes: Midnight Sanctuary (dark glassmorphism), Sage Matcha (earthy green), and Pastel Blossom (soft lavender/pink).

Anyone running on low battery gets to be a little lazier without the guilt—because doing 10% on a bad day is infinitely better than doing 0%.

Demo

Click for Live App

Code

sakshi-db / solace-sanctuary

🦥 An adaptive, low-friction student workspace & life tracker built for low-energy days. Features energy-based task scaling, smart mind dumps, and custom aesthetic themes.

🦥 Solace Sanctuary

The Low-Friction, Adaptive Student Sanctuary & Life Tracker
Built for Hacktoberfest 2026: Week 3 – Slow & Steady (International Sloth Day Theme)

🖤 The Problem: Student Overload & Decision Paralysis

As college students balancing coursework, exams, skill-building, personal habits, and daily life, there are days when we feel completely overwhelmed. Traditional productivity tools often make this worse—they treat every day like a 100% energy day and force you to manually track every single detail, turning task management into yet another chore.

When you're running on low energy, opening a complex planner creates instant burnout and guilt.

🌿 The Solution: What is Solace?

Solace is an anti-overwhelm, low-friction student workspace designed to work with your energy levels, not against them. Instead of pressuring you on low-battery days, Solace adapts to how you feel, automates manual logging, and keeps you moving forward without the friction.

✨ Key Features

🔋…

View on GitHub

How I Built It

Solace is built as a single-page React application powered by Tailwind CSS, Lucide icons, and modern glassmorphism styling. The core intelligence relies on lightweight open-weight text processing models (such as Gemma) to take unstructured "mind dumps" from tired students, parse out completed habits, extract journal notes, and structure tomorrow's action items without requiring manual input forms.

Why Does Open Innovation Matter?

Open innovation and open-weight models make personalized, privacy-first tools accessible to everyone. As a student, having open models means I can run low-friction agents locally or on low-cost infrastructure without worrying about expensive API rate limits or handing private daily diary entries over to closed proprietary tech.

The Lazy Payoff

Solace completely removes the mental chore of "managing a productivity app". Instead of spending 15 minutes organizing lists, setting up priorities, and checking boxes, you dump your thoughts into one box or press one button. It saves at least 20–30 minutes of decision paralysis every morning.

Prize Categories

Best Use of Gemma (or open-weight models)
General Challenge Category: Week 3 – Slow and Steady

Final Thoughts :)

Building Solace helped me reframe how I look at productivity during heavy college weeks. We don't have to be high-performance robots every single day. Taking things slow, keeping momentum steady, and letting smart tools handle the friction is how we build sustainable habits.
If you're a student running on low battery today, take a deep breath. You've got this! 🦥✨

"The Grass-Is-Greener" Procrastination Alarmer

Sakshi Bhosale — Wed, 21 Oct 2026 19:50:13 +0000

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

What I Built

As a computer science student spending hours glued to a screen, my natural habitat is a dark hostel room, VS Code, and an endless stream of black tea. For the "Touch Grass" challenge, I conceptualized an anti-burnout conceptual agent powered by open-weight models that detects when you've been sitting too long and generates sarcastic, low-effort 5-minute outdoor micro-quests to force you outside. It’s built for any student drowning in screen fatigue who needs a gentle, humorous nudge to remember the real world exists.

Demo

Conceptual workflow:
[Screen Inactivity Timer > 3 Hours] ➔ [Local Gemma Model] ➔ [Sarcastic 5-Min Outdoor Quest] ➔ [Student Touches Grass]

Code

sakshi-db / AI-Touch-Grass-Coach

A satirical, deeply relatable AI agent concept designed for developers and college students glued to their desks. Instead of tracking complex outdoor routes, this agent monitors how long you've been sitting in your hostel room or chair. When it detects hours of unbroken coding/scrolling, it uses an open-weight model (like Gemma) to generate a brutally honest, sarcastic, or comforting push notification forcing you to literally "touch grass"—giving you a quick, low-effort 5-minute outdoor micro-quest (e.g., "Go stand under a tree for 3 minutes, drink a glass of water, and look at the sky").

View on GitHub

How I Built It

I explored how open-weight models like Gemma can run locally on a student laptop to parse screen habits and personal activity data completely offline. Keeping the inference local means your burnout data and screen habits stay private on your machine without relying on a costly or privacy-invasive cloud API.

Why Does Open Innovation Matter?

Open innovation matters because you shouldn't have to pay a subscription or hand your private screen-time data over to big tech servers just to be told to step outside. Open-weight models give students the freedom to build privacy-first, custom local tools without API restrictions.

Prize Categories

Best Use of Gemma
General Challenge Category: Week 1 – Touch Grass

Building SecureClinic: A Simple Local PII Redactor Experiment

Sakshi Bhosale — Sun, 18 Oct 2026 14:50:39 +0000

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

What I Built

For this week's challenge, I put together SecureClinic—a small, experimental Python/Streamlit utility designed to scrub Personally Identifiable Information (PII) from text notes locally. It’s a simple concept built for practitioners who want to handle patient notes without leaking data to third-party cloud APIs.

Code & Repo

You can check out the repository and code here:
👉 repo link

How I Built It

Built a quick local web UI using Streamlit.
Set up basic text replacement and file uploading logic to handle raw text files entirely offline.

Why Does Open Innovation Matter?

When building tools that touch sensitive data like medical or health notes, data privacy isn't just an optional feature—it's a hard boundary. Small clinics or independent practitioners cannot risk uploading sensitive records to opaque cloud servers.

Open innovation makes it possible to bring powerful AI capabilities directly to local, air-gapped, or offline environments. A closed commercial API would require sending sensitive data over the internet, whereas open-weight and open-source models let users run tools locally with absolute data ownership and zero cloud leakage.

Just participating, learning, and grabbing that Hacktoberfest swag as I go! Thanks for reading.