DEV Community: shetty sai kumar

EchoMind AI 🎙️ — An AI That Listens, Answers, and Remembers for You

shetty sai kumar — Thu, 22 Oct 2026 19:50:29 +0000

EchoMind AI 🎙️ — An AI That Listens, Answers, and Remembers for You

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

What I Built

Why take notes, search for answers, and organize conversations manually when AI can help?

I built EchoMind AI, a real-time conversation assistant powered by open-weight AI.

Listen → Understand → Answer → Remember

EchoMind AI helps users:

🎙️ Capture microphone and browser-tab audio.
📝 Convert speech into live transcripts.
🧠 Automatically detect questions from conversations.
⚡ Generate contextual answers using Meta Llama 3.1 8B.
💾 Save conversation history and export transcripts.

It's designed for students, developers, online learners, and meeting participants who want to focus on conversations instead of repetitive note-taking.

Demo

🌐 Live App: https://echomind-ai-frontend-mwcb.onrender.com/workspace

Try speaking a question such as "What is Java programming?", then use the question detection feature to generate an AI-powered answer.

Code

💻 GitHub: https://github.com/shettysaikumar20/EchoMind-AI

How I Built It

Frontend: React 19, TypeScript, Vite, Tailwind CSS, Web Audio API, and AudioWorklet.

Backend: Python, FastAPI, WebSockets, SQLAlchemy, and SQLite.

Speech Recognition: ElevenLabs Scribe Realtime (scribe_v2_realtime) processes live audio through a secure FastAPI WebSocket proxy.

Open-Weight AI: Meta Llama 3.1 8B Instruct generates contextual answers through OpenRouter.

Deployment: Render hosts the React frontend and FastAPI backend.

The workflow is straightforward:

Audio → ElevenLabs → Live Transcript → Question Detection → Llama 3.1 → AI Answer

During development, I worked through browser audio permissions, microphone sensitivity, WebSocket disconnections, and speech-recognition accuracy.

Why Does Open Innovation Matter?

Open-weight models make AI development more accessible and flexible.

By using Meta Llama 3.1 8B, EchoMind AI can explore different inference providers and potentially support self-hosted deployment.

Open-source technologies such as React, FastAPI, and SQLAlchemy also make the architecture easier to inspect, modify, and extend.

Instead of building AI models from scratch, I could focus on solving a real problem: reducing repetitive work during conversations.

The Lazy Payoff 🦥

EchoMind AI fits the Slow and Steady theme by reducing the effort needed to capture and revisit conversations.

Instead of manually writing notes, searching for explanations, and organizing answers, users can handle these tasks in one workspace.

Less typing. Less app switching. More listening.

Prize Categories

🏆 Overall Hacktoberfest Week 3 — Slow and Steady
🎙️ Best Use of ElevenLabs — Realtime speech transcription.
☁️ Best Use of Render — Full-stack AI application deployment.
🧠 Best Use of Backboard — AI model orchestration and conversation memory, if integrated into the submitted application.

Final Thoughts

EchoMind AI started with a simple idea: let AI handle the repetitive parts of conversations so people can focus on understanding them.

Building it taught me about real-time audio streaming, open-weight LLMs, WebSockets, and deploying AI-powered applications.

Listen. Understand. Answer. Remember.

🔗 Try EchoMind AI | View Source Code

devchallenge #hf26challenge #opensource #ai

MenuGuard AI 🍽️ — Helping Restaurants Fix Menu Mistakes with Open-Weight AI

shetty sai kumar — Sat, 17 Oct 2026 19:21:51 +0000

MenuGuard AI 🍽️ — Helping Restaurants Fix Menu Mistakes with Open-Weight AI

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

What I Built

Imagine running a restaurant with hundreds of menu items. A duplicate listing, inconsistent price, incorrect dietary label, or misleading description could confuse customers and damage trust.

That's why I built MenuGuard AI, an AI-powered menu auditing platform designed for restaurants, cafés, and cloud kitchens.

It helps business owners identify menu mistakes, review AI-suggested corrections, and export cleaner, more consistent menus.

Key Features

🍽️ Restaurant Dashboard: Track menu items, detected issues, high-priority findings, and audit history.
📂 Menu Management: Add, edit, delete, search, and import menu items using CSV.
🔍 Rule-Based Auditing: Detect duplicate items, pricing conflicts, missing descriptions, and dietary contradictions.
🤖 AI-Powered Auditing: Analyze descriptions, ingredients, potentially misleading claims, and allergen uncertainty.
✅ Human-in-the-Loop Review: Approve, reject, or edit suggested corrections before applying them.
🔄 Before-and-After Comparison: Compare original and corrected menu information.
📊 Export Reports: Download cleaned menus and audit findings in CSV, Markdown, or JSON.

Who Would Pay for It?

MenuGuard AI is designed for independent restaurant owners, small cafés, cloud kitchens, and food businesses that regularly update their menus.

The potential business model is simple: affordable one-time menu audits or a low-cost monthly subscription.

Demo

🚀 Live Application: https://menuguard-ai.onrender.com/

Try the application using the built-in fictional café dataset, run a menu audit, review findings, and explore the correction and export workflow.

Code

💻 GitHub Repository: https://github.com/shettysaikumar20/Menu-Gaurd

The repository includes the frontend, backend, auditing logic, API integration, tests, and deployment configuration.

How I Built It

MenuGuard AI combines deterministic validation with open-weight language model analysis.

Technology Stack:

Frontend: React, Vite, JavaScript
Backend: Node.js, Express
AI Integration: Backboard AI
Open-Weight Model: Meta Llama 3.1 70B Instruct
Model Provider Configuration: Featherless
Testing: Vitest and backend API tests
Deployment: Render
Optional Database: MongoDB

Layer 1 — Deterministic Auditing

The rule engine identifies structural and logical issues such as missing fields, duplicate menu items, conflicting prices, and dietary contradictions.

It also distinguishes legitimate size variants from actual pricing conflicts.

Layer 2 — Open-Weight AI Auditing

Llama 3.1 is integrated through Backboard AI to examine menu descriptions and ingredient information for possible inconsistencies, misleading language, and allergen-related uncertainty.

When credentials are unavailable, the application uses an explicitly labeled mock fallback rather than silently presenting mock findings as live AI results.

Human Review

AI suggestions are not automatically applied.

Restaurant managers can inspect the evidence, compare original and proposed values, approve or reject findings, and explicitly apply accepted changes.

Testing and Verification

During development:

15 backend tests passed.
4 frontend tests passed.
19 automated tests passed in the combined project test run.
Production frontend build completed successfully.
Backend health and API endpoints were tested locally.

The application is configured to deploy its frontend and backend together as one Render web service.

Why Does Open Innovation Matter?

Open innovation is central to MenuGuard AI.

1. Model Flexibility

Using an open-weight model such as Llama 3.1 creates opportunities to experiment with different inference providers and model configurations rather than depending entirely on one proprietary model.

2. Affordable AI for Small Businesses

Restaurant owners need practical tools, not necessarily expensive enterprise software. Open-weight models provide options for building more cost-conscious AI services, although hosted inference still has operating costs.

3. Transparency and Control

The auditing workflow combines understandable deterministic checks with AI analysis. Managers can review evidence and decide which corrections should be applied.

4. Customization

Different restaurants have different cuisines, ingredients, and dietary terminology. An open-weight approach makes it possible to adapt the analysis pipeline and explore specialized models in the future.

5. Open Development

The project's source code is publicly available, making it possible for developers to inspect, improve, and extend the auditing system.

What I Learned

Building MenuGuard AI taught me how to combine traditional software validation with open-weight AI, integrate an external inference service, build human-in-the-loop workflows, test REST APIs, and prepare a full-stack application for cloud deployment.

The biggest lesson was that useful AI products need more than generated answers. They need evidence, clear limitations, and a way for users to remain in control.

Final Thoughts

MenuGuard AI started with a simple idea:

What if a small restaurant could check its entire menu for potential mistakes before customers ever saw them?

By combining rule-based auditing, open-weight AI, and human review, I built a working foundation for an affordable restaurant menu-quality service.

🍽️ MenuGuard AI — Smarter Menus. Fewer Mistakes. Better Business.

Live Demo: https://menuguard-ai.onrender.com/

GitHub: https://github.com/shettysaikumar20/Menu-Gaurd

Thanks to DEV, MLH, and the Hacktoberfest community for encouraging developers to build useful products with open AI!

OFFLOOP AI: ARISE — Turn Real Life Into an RPG and Break the Doomscrolling Cycle

shetty sai kumar — Sun, 11 Oct 2026 17:32:07 +0000

OFFLOOP AI: ARISE ⚔️ — Stop Scrolling, Start Leveling Up!

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

What I Built

OFFLOOP AI: ARISE is an AI-powered, RPG-inspired application that turns real-world activities into exciting quests.

Instead of endless scrolling, users become Hunters, complete challenges, earn XP, improve their stats, and level up in real life.

Key Features:

⚔️ Hunter Awakening and Rank Progression
🌳 AI-generated outdoor exploration quests
🏆 XP, achievements, and daily challenges
💪 Character attributes and progression
🎮 Immersive fantasy RPG interface

Mission: Make going outside as exciting as playing a game!

Demo

🌐 Live Website

🎬 YouTube Demo

Code

💻 GitHub Repository

How I Built It

Frontend: React, Vite, JavaScript, CSS
Backend: Python, FastAPI
Open-Weight AI: Google Gemma 2 (gemma2:2b) via Ollama
Location Data: OpenStreetMap + Overpass API
Deployment: Render

The locally tested Gemma 2 backend generates location-based outdoor quests with creative narratives and clues.

Example AI-generated quest:

🌳 Location: Cubbon Park Botanical Garden

⚔️ Quest: Nature's Symphony: A Botanical Journey

🤖 Model: Gemma 2 (2B)

✅ Result: Successfully generated a narrative and three exploration clues.

The backend also supports hosted Gemma and deterministic fallback generation. The live website currently runs its core RPG features in the browser; the Gemma backend runs separately.

Why Does Open Innovation Matter?

Open-weight AI makes intelligent applications more accessible, affordable, and customizable.

Using Gemma 2 and Ollama, I could develop and test AI-powered quest generation locally without depending entirely on proprietary cloud APIs.

Open innovation empowers students and developers to experiment, collaborate, and build meaningful solutions.

My Agent Session

I used AI-assisted development tools for coding, debugging, testing, and documentation.

Final Thoughts

Technology should inspire us to experience the real world, not just stare at screens.

⚔️ OFFLOOP AI: ARISE — Your Real Life Is Your Dungeon.

Stop Scrolling. Start Leveling Up!

[Boost]

shetty sai kumar — Sun, 04 Oct 2026 22:27:14 +0000

shetty sai kumar

Oct 4

I Built My Mother an AI Study Companion That Still Works Without the Internet

#devchallenge #weekendchallenge #hf26challenge #opensource

5 min read

I Built My Mother an AI Study Companion That Still Works Without the Internet

shetty sai kumar — Sun, 04 Oct 2026 22:24:36 +0000

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

What I Built

For the past three months, my mother has been learning 10 English words every day.

Telugu is our native language, and she wants to become more comfortable understanding and speaking English. Her consistency wasn't the problem — retention was.

She could learn ten words today, but weeks later, when I used one of those words in conversation, she would sometimes struggle to remember what it meant or how it was used.

That gave me the idea for FriendForge — a Personal AI Study Companion.

I didn't want to build another chatbot that simply answers questions. I wanted something that could help her:

learn words through simple explanations and examples
quiz herself
revise what she struggled with
learn from uploaded notes
listen to explanations
continue learning even when internet connectivity isn't available

That last requirement became one of the most important parts of FriendForge.

My mother sometimes travels to villages and places where internet connectivity can be unreliable. So FriendForge has two different ways to learn:

🌐 Online Study — Llama through Backboard, with RAG, study memory, quizzes, revision, and ElevenLabs voice.

💻 Local AI — Gemma 3 4B running locally through Ollama, allowing general study conversations without an internet connection once the model is installed.

The idea behind FriendForge became:

RAG remembers the material. Memory remembers the learner.

Demo

🎥 Video Demo

🌐 Live Application

https://friendforge-gbqq.onrender.com

🔑 Demo access: HACKTOBERFEST2026!

The hosted version demonstrates FriendForge's Online Study experience. Local AI runs through Ollama on the user's own computer, so I demonstrate that separately in the video.

Code

The complete project is available on GitHub:

shettysaikumar20 / FriendForge

AI study companion with RAG, learner memory, local Gemma inference, and voice explanations.

⚡ FriendForge

An AI Learning Companion — Learn. Practice. Remember. Revise.

RAG remembers the material. Memory remembers the learner.

Online Study ☁️ · Document RAG 📚 · Study Memory 🧠 · Quizzes 📝 · Voice 🔊 · Local AI 💻

✨ What is FriendForge?

FriendForge is a personal AI learning companion built for the Hacktoberfest 2026 — Build for a Friend Challenge.

Instead of acting as just another chatbot, FriendForge combines learning material, contextual explanations, active recall, quizzes, revision, learner memory, voice accessibility, and optional local AI.

A learner can upload study material, ask questions about it, simplify difficult concepts, test understanding, revisit weak areas, and switch to a locally running open-weight model when hosted AI is not appropriate or internet connectivity is unavailable.

The goal isn't only to answer a question. It's to help the learner understand it, test it, and come back to what they haven't mastered…

View on GitHub

FriendForge uses a React/Vite frontend with a Node.js/Express backend.

How I Built It

FriendForge has two intentionally separate AI paths.

🌐 Online Study

The online workflow is approximately:

React → Express → Backboard → Llama

Backboard provides the infrastructure for the online AI experience, including document retrieval and learner memory.

Users can upload PDF or TXT study material and ask questions based on those documents.

FriendForge also provides Quiz Me and Revise workflows so learning doesn't stop after receiving an explanation.

For voice learning, I integrated ElevenLabs text-to-speech. Voice generation happens only when the learner explicitly presses Listen.

💻 Local AI

The local workflow is:

React → Express → Ollama → Gemma 3 4B

This was particularly important to me.

Once Ollama and Gemma are installed, the model inference happens directly on the laptop. FriendForge doesn't silently fall back to a cloud model when Local AI is selected.

Local mode deliberately has a smaller feature set. It doesn't pretend to have Backboard document RAG, persistent cloud memory, or ElevenLabs voice.

It is simply a local study companion.

Building for an 8 GB Laptop

I didn't develop the offline feature on a powerful AI workstation.

I tested Gemma 3 4B on my 8 GB RAM laptop using CPU inference.

That exposed a very real trade-off.

The hosted model responds quickly. Local Gemma can take considerably longer — sometimes tens of seconds depending on model loading and the prompt.

But it works without internet access.

For this project, that trade-off was worth making visible instead of hiding it.

Things Broke Along the Way

One of my favorite parts of this project was discovering that making an AI feature work is different from making it reliable.

During RAG testing, I found a case where the model correctly recognized that information wasn't present in the uploaded notes, but then continued by supplying outside knowledge anyway.

That wasn't the behavior I wanted for a question explicitly asking about uploaded material.

I tightened the retrieval boundaries and added tests around that behavior.

I found a similar problem with local Gemma. Initially, it could imply that it had access to uploaded notes even though Local AI intentionally doesn't have access to Backboard RAG.

So I strengthened the Local AI capability boundaries and tested them as well.

Even production deployment gave me another surprise.

ElevenLabs successfully generated audio, but the browser refused to play it because my Content Security Policy didn't permit the generated Blob URL.

The fix was intentionally narrow:

media-src 'self' blob:;

Then I added a regression test instead of weakening the rest of the CSP.

These failures ended up teaching me as much as implementing the original features.

Why Does Open Innovation Matter?

For FriendForge, open innovation isn't an abstract argument.

It's the reason part of this application can continue working when the internet disappears.

Running Gemma 3 4B locally through Ollama means the core local study conversation doesn't require sending every question to a hosted AI service.

It also gives me freedom to experiment with the model and local inference architecture rather than making the entire learning experience dependent on one remote API.

There is a cost.

On my hardware, local inference is significantly slower than the hosted experience.

But when I disconnected Wi-Fi and watched FriendForge continue generating an answer, the advantage became very clear to me.

Speed wasn't the advantage. Independence was.

Open-weight models make it possible to build experiences where cloud AI and local AI don't have to compete. They can solve different parts of the same problem.

FriendForge uses the cloud when connectivity and richer features are available, and provides a smaller local experience when independence matters more.

The Most Important Test

Eventually, I gave FriendForge to the person I actually built it for.

My mother tried it.

The difference between the two modes was immediately noticeable. Online Llama was fast, while local Gemma took more time to respond on my laptop.

But the part that surprised her was that the local AI could still answer when there was no internet connection.

She told me she was proud of me and that the project would be useful to her.

That mattered more to me than another benchmark.

FriendForge isn't something I built by imagining what a hypothetical user might want.

It came from watching someone I love consistently try to learn something, noticing where technology could help, and then letting her actually use what I built.

What FriendForge Can't Do Yet

I also want to be transparent about its limitations.

Local AI currently provides general study conversations. It doesn't have the document RAG, persistent Backboard study memory, or ElevenLabs voice available in Online Study.

Local inference is also considerably slower on my CPU-only 8 GB laptop.

And the hosted Render deployment cannot access Ollama running on somebody's laptop. Local AI is therefore used through the locally launched FriendForge application.

Those aren't features I want to disguise. They're boundaries that make the two modes understandable and predictable.

🤖 My Agent Sessions

I used an AI-assisted development workflow throughout FriendForge and preserved the development process with DevRelay.

1. Building FriendForge & the RAG Foundation

This session covers the initial architecture, React/Vite frontend, Express backend, Backboard integration, and the development of the document-based RAG workflow.

https://dev.to/agent_sessions/friendforge-open-weight-ai-study-companion-full-development-rag-journey-ddiqhi

2. Reliability, Local AI & Final Refinement

This session captures the later engineering and reliability work as FriendForge evolved into the final application.

https://dev.to/agent_sessions/friendforge-open-weight-ai-study-companion-full-build-reliability-journey-ly6gby

Technology

Gemma 3 4B — open-weight local model
Ollama — local inference
Backboard — online AI, RAG, and memory
Llama — online language model
ElevenLabs — text-to-speech
React + Vite — frontend
Node.js + Express — backend
Render — deployment

Prize Categories

I'm submitting FriendForge for:

Overall Challenge
Gemma
Backboard
ElevenLabs
Render

I started FriendForge because my mother was learning ten English words every day.

I finished it after watching her use something I had built specifically for the way she learns — including seeing her surprise when the AI continued responding without an internet connection.

Speed wasn't the advantage. Independence was.

And for me, building something genuinely useful for someone I love is exactly what Build for a Friend meant.