Title:

Dialogix – Conversational Gateway

Workplace conversations like asking for a raise or handling conflict can shape a career, but most people never get a safe space to practice them before the stakes are real. We built Dialogix to close that gap. The app runs on React Native and Expo for the frontend, with a Node.js TypeScript backend using Express and MongoDB. AI coaching is powered by OpenAI GPT-4o-mini and a Pinecone vector database. Every suggestion the model makes is grounded in Dr. Dennis Rebelo's three storytelling frameworks: Peak Storytelling, Storytelling Flow, and Behavior-Over-Time. Together these map user responses to nine leadership archetypes across past, current, and future layers. Each session runs a retrieval-augmented generation pipeline. The app embeds the user's input, pulls relevant past conversations and LMS strategy content from Pinecone, then generates four archetype-based suggestions per turn. Once the session ends, it scores the narrative and tracks growth with weekly and monthly progress metrics. We tested auth flows, scoped data access, rate limiting, and cross-user isolation across fifteen-plus endpoints, and everything held up. The result is a working platform where people can rehearse difficult conversations and get real-time feedback rooted in communication theory rather than generic advice.

Name

Comment*

Comments are viewable only by submitter

Captcha*