Abel Mendez

One-Arm Parent is a voice-first web app for caregivers who are holding a baby and can't touch their phone. It solves a universal problem: every new parent ends up exhausted at 3 AM, one hand on the baby, unable to type, tap, or navigate a traditional app. One-Arm Parent lets you do everything by voice. Say "start sleep routine" and the app guides you step-by-step through a calming bedtime sequence with ambient sounds like white noise, rain, or heartbeat. Say "log feeding" or "dirty diaper" and the app records it instantly with no forms or keyboards. Say "what happened today" and it reads a spoken summary of the baby's day so you can hand off context to the other caregiver without writing anything down. The app even whispers its responses during sleep routines to avoid waking the baby, and includes soft voice reminders so you never miss the next feeding or nap window. The entire audio experience is powered by ElevenLabs: Conversational AI understands commands and responds naturally in a warm, calm tone, Text-to-Speech delivers all confirmations, guided steps, whispered prompts, and summaries aloud, and every pre-generated asset including lullabies, ambient sounds, soothing textures, confirmation chimes, reminder tones, and fallback phrases was created using ElevenLabs' audio generation capabilities. The app was built end-to-end with Cursor, which accelerated everything from scaffolding the Next.js architecture and designing the XState state machine to wiring the ElevenLabs SDK and iterating on voice command parsing. The design is intentionally voice-first, not a traditional UI with voice bolted on, with graceful error recovery, closed intents for reliability, and always-visible mic and audio state so the caregiver never has to stare at a screen. Built with Next.js 16, React 19, TypeScript, XState, Howler.js, Tailwind CSS, and the ElevenLabs SDK.

Submitted 14 May 2026

We completely redesigned the official FIFA World Cup 2026 Hospitality website (https://fifaworldcup26.hospitality.fifa.com/us/en) as a dark, cinematic, editorial experience and added something the original doesn't have: a voice concierge powered by ElevenLabs that lets you talk to the website. The Problem: The 2026 World Cup is the most complex in history. Three host countries, 16 venues, 104 matches, 48 teams. Navigating this information through traditional filters and tables is overwhelming, especially for fans planning cross-border trips. Our Solution: A voice-first AI concierge that knows every match, venue, hospitality package, and team schedule. Ask "Where does Argentina play?" and it answers instantly, then navigates you there. Ask "What matches are in Mexico City between June 14-18?" and it cross-references everything for you. The agent doesn't just answer. It acts on the UI: scrolling to sections, navigating between pages, and highlighting relevant content. Design: We replaced the corporate light theme with a layered dark luxury aesthetic. Radial glows, pitch blueprint SVG overlays, animated ball trajectory paths, glassmorphism panels, 3D Three.js hero scene, editorial section numbering, and a tri-accent color system (Cyan / Gold / Pitch Green). Every section feels like a premium sports magazine. ElevenLabs Integration: The Conversational AI SDK powers a persistent floating voice dock across all pages. We feed the full tournament data (104 matches, 48 teams, 16 venues, pricing) into a RAG knowledge base. The agent uses client-side tool calling to navigate the site, making voice a first-class interaction method, not a gimmick. Built with: v0, Next.js 16, React 19, Three.js, Tailwind CSS v4, ElevenLabs Conversational AI SDK

Atlas Cipher is a voice-first detective game where players receive missions from a handler, review case briefings, travel across a 3D globe, interrogate witnesses, collect clues, compare suspect dossiers, and ultimately decide who to arrest. Each case is structured as a focused but complete investigation loop: players manage limited time and travel resources, revisit their notebook, track evidence across cities, and piece together route clues, suspect clues, and corroborating details before issuing a final warrant. We built Atlas Cipher in Zed and used ElevenLabs as a core part of the experience, not just as an audio layer. ElevenLabs powers the live conversational interactions with the handler and witnesses, character voice design, narration, ambient sound, mission music, and supporting sound effects that shape the emotional pacing of the investigation. Our goal was to create a smaller, polished game experience where audio drives immersion, tension, and personality, making every conversation, city visit, and final accusation feel more immediate and memorable.

Reel&Ink is an AI-powered animated story studio that turns a single text prompt into a fully produced animated video. You describe a story idea, and the app generates everything: visual style, characters with unique AI-designed voices, locations with background art, a structured screenplay, full audio production (narration, dialogue, music, sound effects), and a final animated video, all playable in the browser. The entire project was built using Kiro's spec-driven development. Requirements, technical design, and implementation tasks were defined as specs upfront, and Kiro's AI agent implemented them systematically. Steering files kept every generated file consistent, hooks automated type syncing and linting, and MCP servers (including Firecrawl for web access) extended the agent's capabilities during development. A custom HyperFrames Power was built using Kiro's Power Builder to give the agent deep knowledge of the video composition framework. ElevenLabs powers every sound in the app: the Voice Design API creates unique character voices from text descriptions, the TTS API generates expressive narration and dialogue with word-level timestamps, and the Music and Sound Effects APIs compose original scores and ambient audio per scene. Those word-level timestamps are what drive the final video, syncing subtitles, character animations, and scene transitions to the spoken audio through HyperFrames HTML compositions with GSAP timelines.

SoundCharades is an AI-powered audio guessing game where every sound effect and original song is generated from scratch by ElevenLabs. Players listen to layered soundscapes and AI-composed riddle songs to guess movies, video games, countries, food, and more. The game's difficulty engine is built entirely on turbopuffer, using cosine similarity across 768-dimensional embeddings to dynamically select answer options based on semantic proximity. The closer the concepts are in vector space, the harder it is to tell them apart. Players can also generate custom quizzes on any topic in minutes, with new concepts instantly queryable in turbopuffer the moment they're created. No manual curation, no static databases, just vectors, sound, and AI.

ToughTalk AI - Practice any difficult conversation before it happens, and hear yourself nailing it. Pick a scenario, an angry customer, a salary negotiation, an aggressive interview, and start talking. The app creates a unique AI character with its own personality, tactics, and a custom-designed voice through ElevenLabs. You practice live, by voice, directly from your browser. The character pushes back, interrupts, and resists just like a real person would. After the session, you get a coaching report with a performance score, your strengths backed by exact quotes, and specific phrases you can use next time. Then comes the magic. Clone your voice with a 30-second recording, and ToughTalk rebuilds the entire conversation as an ideal replay — two ElevenLabs agents talking to each other using the character's designed voice and YOUR cloned voice. You literally hear yourself handling it perfectly. The app uses five ElevenLabs capabilities deeply: Voice Design to create character voices from descriptions, Conversational AI for real-time practice via signed URLs, Instant Voice Cloning from a short sample, Text-to-Speech for replay generation, and a dual-agent backend orchestration where two ElevenLabs agents converse autonomously to produce the ideal replay. Built and deployed on Replit, using Replit's integrated PostgreSQL and object storage.

Dentists spend a significant portion of every appointment on documentation, recording findings tooth by tooth while examining a patient is slow, error-prone, and breaks clinical focus, whether they work with an assistant or alone. Aural solves this with real-time voice-controlled dental charting. The dentist speaks naturally while working, the odontogram updates instantly, the agent confirms every entry out loud, answers questions about the patient's history mid-appointment, and generates the full clinical report automatically at session end. ElevenLabs and Cloudflare are both central to how this works. We used three ElevenLabs services — Conversational AI for agent mode with tool use, Scribe for real-time speech recognition, and TTS for spoken confirmations and summaries — running on eight Cloudflare services including Workers AI for dental NLU, Durable Objects for live WebSocket session state, D1 for patient history, and R2 for report storage. One real clinical problem, solved end to end at the edge.

Pulse is an AI-powered live radio station that broadcasts 24/7 news on AI and startups. It uses ElevenLabs Conversational AI to power an autonomous presenter with editorial personality, AI co-hosts for debate segments, and expert guests who break down stories in depth. Listeners can call in live and interrupt naturally with barge-in support. Firecrawl continuously scrapes news sources, feeding a live pipeline of breaking stories with confidence levels, triggering automatic breaking news interruptions when stories explode.