STANECT CASE STUDY

Quick Facts

Role:

Product Engineer

Ownership:

Built the entire product end-to-end

Users:

200+ beta users

Core Stack:

Next.jsReactTypeScriptTailwind CSSFramer MotionSupabaseVapiOpenAI

Focus:

Reliable onboarding, scalable architecture, AI conversation flow

Overview

Stanect is an AI-powered voice companion designed to help users practice conversations and improve communication confidence. I built the product end-to-end, including the landing experience, onboarding system, companion creation flow, real-time conversation interface, backend architecture, and subscription infrastructure.

The Problem

Many people want to improve their communication skills but lack safe environments to practice conversations. I explored whether an AI voice companion could create a low-pressure space where users can experiment, practice responses, and build confidence speaking to others.

Building the Product

Stanect was developed as a full-stack SaaS application. I designed the entire user journey—from landing page to first conversation—while building the frontend architecture, backend data systems, onboarding flows, and subscription infrastructure needed to support real users.

Key Features

Companion creation flow

Preview-driven configuration

Real-time conversation sessions

Profile management

Subscription model for premium access

Architecture

The application was built using a modern full-stack architecture.

Frontend:

Next.js

React

TypeScript

Tailwind CSS

Framer Motion

State and data management:

Zustand

TanStack Query

Zod

Backend infrastructure:

Supabase

Clerk

Voice and AI processing:

Vapi for voice processing

OpenAI for conversation generation

This architecture allowed the application to remain scalable while keeping the development workflow efficient.

Key Engineering Decisions

Persistent onboarding state to prevent progress loss

Multi-layer validation to guarantee safe companion creation

Modular component architecture for long-term maintainability

Engineering Challenges

One challenge was balancing strict validation with a smooth onboarding experience. Too much validation can block users, while too little risks invalid data entering the system.

Outcome

Stanect evolved into a complete SaaS product with structured onboarding, AI-driven conversation sessions, profile management, and subscription flows.

Product Thinking

While building Stanect, I focused on minimizing time to first conversation, providing clear user feedback during onboarding, and structuring the codebase so the product could scale as features grow.

Outcomes

200+ beta users supported

End-to-end onboarding to conversation flow

Layered validation across UI and server

Scalable component architecture

Demo Assets

Onboarding step flow • Companion creation preview • Conversation interface • User profile dashboard

Tech Stack

Next.js (App Router)ReactTypeScriptZustandTanStack QueryZodSupabaseClerkVapiOpenAITailwind CSSFramer Motion