EmotiSync

EmotiSync

stealth

Personal AI voice companion for emotional well-being.

Timeline

3 months, October 2024 - November 2025

Role

Founding Engineer

Stack

Next.jsTypeScriptTailwind CSSOpenAI APIHume AI APIPostgreSQLSupabaseDrizzle ORMUpstashZustandZodBunRedis

Overview

Skip writing, just talk naturally. Your AI companion transforms your daily voice conversations into personalized insights and recommendations, helping you find clarity and growth in minutes. Empathetic voice AI: Express yourself through natural voice conversations with AI who listens and responds with empathy. Mood-boosting recommendations: Receive personalized actionable recommendations that actually work for you. Automatically generate journal summaries that track your reflection progress - zero writing required. Get personalized insights tailored to your unique mood. Customize your AI avatar: Customize your AI avatar to your unique personality.

Problem

Most mental wellness apps are passive — they track mood but don't actually engage with you. People struggling emotionally need something that listens and responds in real time.

What I Built

Built and shipped a full production AI voice companion — real-time empathetic voice conversations powered by Hume AI, automated journal summaries via OpenAI API, mood insights, auth, payments via Stripe, and rate limiting. Sole engineer on a live product with real users.

Challenges

Handling real-time voice AI reliably in production was technically demanding — latency, interruptions, and emotional tone detection all needed careful tuning. Building a product that felt genuinely caring rather than robotic required iterating heavily on the conversation flow.

Outcome

Shipped to production at emotisync.xyz. Acquired early users during testing phase. Full production architecture including voice AI, payments, auth, and database — built and maintained solo.

EmotiSync

Team

  • Samuel Danso