Love Bloomer: From Concept to Completion
Project Overview
Brainstorming the Original Idea
The idea for Love Bloomer started from a common issue many people face on modern dating apps: finding the right words during important moments in a conversation. While AI had already made strides in other industries, I wanted to create something unique. I envisioned an AI-powered platform that would assist users in romantic conversations by providing tailored responses. This way, users could overcome awkward silences or difficult questions and increase their chances of making a lasting connection.
Early Stages: Tech Stack and Design
After brainstorming, I moved on to the technical aspects. The frontend was built using Angular due to its component-based architecture, making it easy to scale. Nest.js was chosen for the backend because of its TypeScript compatibility, perfect for API integration. MongoDB was used for flexible data handling, and secure authentication was provided by Auth0. I also realized that modern, clean designs would play a key role in the app's appeal. I started designing interfaces that would be engaging and intuitive.
Landing Page
The landing page was designed to introduce users to the app's purpose clearly and quickly. The prominent call-to-action invites users to log in and start the experience, while the visual aesthetics focus on a sleek, modern design with a welcoming tone.
Developing Key Features
Once the designs were set, development began. The most critical part of the project was the conversation feature. I built a system where users could enter their chat dialogues, and the AI would suggest appropriate responses, helping them navigate the conversation.
User Profile Input Page
In this screen, users input their name and the name of their chat partner. This helps create a more personalized interaction with the AI. Users can also add persona descriptions to guide the AI in generating responses that fit the personalities involved in the conversation.
New Description Input
Users can create detailed descriptions for themselves or their partner, enhancing the AI’s contextual understanding. This was a crucial feature in tailoring AI-generated responses to the specific dynamics of each conversation.
Conversation Interface
The conversation interface shows the dialogue between the user and their chat partner. Users can request the AI to suggest replies, rewrite messages, or add more responses, allowing for a fluid and interactive chatting experience.
Conversation Suggestions
When the user isn't sure how to respond, the AI steps in by suggesting multiple conversational lines. This feature is particularly helpful when navigating tricky or important moments in the dialogue.
Challenges and Solutions
One of the biggest challenges was ensuring the AI-generated responses felt natural and personal. To solve this, I refined the AI by allowing users to provide detailed contextual information, such as descriptions and dialogue history, which helped the AI produce more relevant suggestions.
Conversation Dashboard
The dashboard provides a clean and organized overview of ongoing conversations. Users can track the messages they've exchanged and request new suggestions from the AI when necessary.
Credit Management and Payment Integration
Since the AI-based responses require processing power, users are allocated credits. If they run out of credits, a notification prompts them to top-up their account to continue using the service. Stripe was integrated for seamless payment processing.
Future Plans
The initial launch of Love Bloomer has been a success, but there’s much more in store. I plan to implement even more predefined conversation templates, enhance the AI’s understanding of nuanced dialogues, and add new features like conversation analytics, which would help users track their progress and become more adept at online conversations.