gowtham jayavarapu

VoiceChef is a fully voice-controlled cooking assistant designed for moments when your hands aren’t free — mixing dough, cooking with oil, washing vegetables, or multitasking in the kitchen. Most cooking apps still rely heavily on touch interaction, which becomes frustrating and messy during real cooking. VoiceChef solves this by creating a completely hands-free cooking experience powered by voice AI. Users can: 🎙️ Navigate recipes entirely with voice commands 📸 Snap a photo of a recipe and instantly extract steps using AI 🗣️ Ask cooking questions naturally while cooking ⏱️ Set timers hands-free 👨‍🍳 Cook alongside different AI chef personalities like “Nani”, “Drill Sergeant”, “Zen Master”, and “Hype Man” VoiceChef uses ElevenLabs Speech-to-Text to understand user voice commands in real time and ElevenLabs Text-to-Speech to bring the chef personalities to life with natural, expressive voices. The app was built using Cursor as the primary AI-first development environment, helping rapidly scaffold, iterate, debug, and refine the entire experience during the hackathon. The goal was to create a cooking experience that feels conversational, immersive, and effortless — where users can simply talk and cook without ever touching a keyboard or screen.

Submitted 14 May 2026

Audible has over a million audiobooks and world-class narrators. But their website makes you feel nothing. No atmosphere, no emotion, no way to discover what to listen to based on how you actually feel right now. This redesign started with one question — what if the website actually matched the product? The problems I identified: Discovery is broken. Audible asks you what genre you want. But nobody thinks in genres. They think in feelings. I'm stressed. I can't sleep. I want to escape. So I replaced genre dropdowns with 8 emotional states. You tell it where you are, it finds what you need. Book pages tell you nothing personal. They show you a price and a runtime. They don't tell you if this book is right for you, right now. So I added an AI that has actually read the book. You ask it anything — it answers out loud. The website is completely silent. This is a product sold entirely on the power of voice and narration. So every interaction that matters — recommendations, book answers, genre teasers — speaks back to you through ElevenLabs. Voice in, voice out. Tech Stack: — Google Stitch for visual design direction — v0 by Vercel for UI build — Claude for architecture and design decisions — Groq for real-time AI responses — ElevenLabs for all voice output

SUSPECT is an AI-powered interrogation game where you sit across from Detective Harlow and answer his questions with your actual voice. There are three cases — a murder, a missing person, and a suspicious accident. Each one starts with a cinematic briefing that puts you in the situation. Then the interrogation begins. You speak your answers out loud. The game transcribes them in real time. The detective reads exactly what you said and responds — not from a script, but generated fresh each time based on your specific words. If you contradict yourself, he catches it. If you stay silent, he notes it. A 45-second timer ticks audibly throughout your answer window. As pressure escalates across 6 questions, the room visually closes in — the vignette darkens, the light turns red, the detective's voice gets colder. At pressure level 5 the atmosphere is genuinely suffocating. The voice cloning feature is the centerpiece: after you've spoken enough, the game silently clones your voice using ElevenLabs' API. At a critical moment in the interrogation, it plays your own words back at you — in your own voice — as evidence. The disconnect between what you said then and what you're saying now is what breaks most players. At the end: CONFESSION, RELEASED, or NO COMMENT. The AI judge reads your full transcript and decides. Built entirely in Zed — the AI-native code editor. Zed's built-in AI assistant was used throughout development to scaffold components, write the interrogation logic, and debug the real-time audio pipeline. The speed of Zed's AI panel made it possible to ship this in 48 hours. ElevenLabs APIs used: — Text-to-speech (Detective Harlow's voice, generated per question) — Voice cloning (suspect's voice captured and replicated mid-game) — Sound effects generation (ticking clock, ambient drone, slam SFX) — Speech-to-text (transcribing the player's spoken answers) The goal was to make a game that makes you feel something. Most players confess.

Talkify is an AI-powered app that transforms everyday objects into interactive personalities. By simply pointing your camera at something like a toy, a book, or any object around you, Talkify generates a unique identity for it — complete with a name, personality traits, and even a backstory. What makes it special is that these objects don’t just exist visually; they can actually talk, respond, and even sing using expressive, human-like voices. Built using Kiro and powered by hyper-real voice technology from ElevenLabs, Talkify turns ordinary moments into engaging, interactive experiences. It reimagines how we interact with the world around us by giving everything a voice.

Objects That Sing is an AI-powered music app that turns your surroundings into a personalized motivational song. What it does — the full flow: You photograph anything around you — your desk, your kitchen, your gym, your bedroom. Any photo works. AI scans the photo and detects 3–5 objects — using Groq's Llama 4 vision model. It picks the most interesting, character-rich objects (not background noise). Each object gets a unique personality — fetched from Turbopuffer, a vector database seeded with personality profiles for hundreds of objects. A coffee mug has different energy than a dumbbell. You set the vibe — four choices: Your goal right now (free text — "I need to stop procrastinating", "I need to call my mom") Music genre (Hip Hop, R&B, Pop, Rock, Gospel, Lo-fi, Bollywood, K-pop, Electronic, Jazz, Reggaeton, Country) Intensity (Gentle Nudge / Medium Fire / Full Beast Mode) Language (English, Hindi, Tamil, Telugu, Spanish, French, Portuguese, Arabic, Korean, Japanese) Groq writes the lyrics — using Llama 3.3 70B. Each object gets a verse where it speaks directly to you in first person, mid-conversation, with attitude. The object has a memory of your habits, an opinion about you specifically, and ends every verse with one direct demand. The chorus is all objects together at maximum energy. ElevenLabs sings the entire song — a full 90-second sung song with real vocals generated via the /v1/music API. Not text-to-speech — actual music with singing. You listen in the Player — with your photo as cover art, scrolling lyrics synced to playback, genre badge, object tags, share button, and a download option. Library saves all your songs — everything you generate is saved to your profile for replay anytime. Tech stack: Layer What's used Frontend React + Vite, Tailwind CSS, Framer Motion Backend Express (Node.js) AI Vision Groq — Llama 4 Scout 17B Lyrics Groq — Llama 3.3 70B Versatile Music ElevenLabs /v1/music (sung vocals) Personalities Turbopuffer vector database Database Replit PostgreSQL (via Drizzle ORM) Sessions Cookie-based (no login required) Storage Persistent file storage for photos + MP3s Design: Mobile-first — works perfectly on any phone browser Desktop — shows the app inside a realistic iPhone 15 Pro frame with an animated dark space background (aurora blobs, floating music particles, glassmorphism side panel) iOS light-mode inside the phone — white cards, iOS blue accents, real shadows, native feel What makes it different: The objects aren't motivational speakers. They're characters with attitude. A cable machine that's furious you walked past it. A carrot that's personally offended you keep choosing pizza. A notebook that remembers you haven't opened it in three days. The song is written specifically for you, about your specific goal, using only the objects actually in your photo — nothing generic.

CastVoice is a cinematic story-to-audio-drama platform that transforms written stories into fully produced audio experiences. Each character is assigned a distinct voice, scenes include contextual sound effects, and the final output is a fully mixed audio drama. Characters can be cast using AI-designed voices, cloned voices, or by inviting others to contribute voice recordings. What problem does it solve? Producing high-quality audio storytelling traditionally requires voice actors, sound design, and manual editing across multiple tools. Existing AI tools focus primarily on text or single-voice output and do not support multi-character storytelling or end-to-end production. CastVoice addresses this by automating script structuring, voice generation, sound design, and audio mixing, while enabling collaborative storytelling through shared voice participation. How does it use ElevenLabs and this week’s sponsor? ElevenLabs powers the core audio pipeline, including text-to-speech generation with emotion-aware delivery, voice cloning, and contextual sound effect generation. Each character’s dialogue and scene audio are generated and combined into a single, cohesive audio output. Replit is used for building, running, and deploying the full-stack application, enabling rapid development and integration throughout the project.

Vibe Echo is a real-time social game where you don’t talk — you vibe. Two players are instantly matched and dropped into a shared session where the only way to communicate is through expressive reactions. Each reaction triggers unique AI-generated character sounds, turning every interaction into a playful, unpredictable exchange. No text. No voice chat. Just pure energy.

Newzzy is a personal AI news companion that lets you talk to the world's latest news in real time. Ask anything — it searches the live web instantly using Firecrawl Search and speaks back a natural, conversational answer through ElevenLabs Agents, while showing real source cards on screen simultaneously. Not satisfied with the headlines? Jump into Debate Arena, pick your tone (calm, aggressive, or roasting 🔥), and argue live with an AI that fights back using actual Firecrawl-sourced facts. One voice. The whole world. Right now url:https://newscompanion.vercel.app/