LifeWork — Memorial Chatbot
Talk to Friend — a memorial chatbot that keeps someone's memory alive through conversation
Problem
Memorial conversations needed a chatbot that could reflect a person's tone and improve when it could not answer.
My contribution
Built the memorial chatbot with persona editor, learning queue, mood tracking, and admin dashboard.
Technology
NLP · Real-time chat · Self-learning pipeline · Persona engine
Result
Working prototype with self-learning pipeline — admin can refine unmatched queries over time.
Overview
"Talk to Friend" — A memorial chatbot that keeps someone's memory alive through conversation. Built as a personal passion project to explore how NLP and personality modeling can create meaningful, lasting connections.
Key features
- Persona editor to customize the chatbot's personality and tone
- Admin dashboard with conversation stats and match rates
- Learning queue for unmatched queries — the bot improves over time
- Mood tracking and memory storage across sessions
- Dark-themed, elegant UI designed for warmth and comfort
Technical focus
NLP-based chatbot with a self-learning pipeline, persona configuration, and real-time chat. The system learns from unanswered questions and lets an admin refine responses over time, so the bot becomes more accurate and personal with use.
Status
Active work in progress — case study will grow as the project evolves.
Gallery
