Pro_Compose

Pro_Compose is a modern, professional AI-powered email generator to quickly draft high-converting, professional emails based on your specific requirements like tone, formality, and key talking points.

Comment

Inspiration

Drafting professional, context-appropriate emails is often time-consuming and tedious. Sometimes it’s difficult to strike the perfect balance between professional and friendly, or to ensure that all necessary talking points are covered clearly without sounding robotic. We wanted to build a tool that removes this friction by harnessing the speed and intelligence of Google's Gemini 2.5 Flash Lite model. The goal was to create a modern, beautiful wrapper that helps professionals, students, and businesses communicate more effectively and efficiently.

What it does

Pro_Compose is a sleek, AI-powered email generator designed to instantly draft high-converting, polished emails. Users simply input the core bullet points they need to communicate, select their desired tone (e.g., Professional, Friendly, Urgent, or Persuasive), and adjust a formality slider (from 1-10).

With a single click, Pro_Compose processes these parameters and generates a fully fleshed-out email draft that strictly adheres to the requested tone and includes all required details. Users can then copy the draft to their clipboard instantly to use in their preferred email client.

How we built it

We adopted a modern full-stack architecture to ensure the app is fast, scalable, and a joy to use:

  • Frontend: Built with React 18 and TypeScript via Vite for lightning-fast development. We used Tailwind CSS for responsive, utility-first styling and Framer Motion to add fluid micro-animations that give the app a premium, dynamic feel. Icons were provided by Lucide React.
  • Backend: Powered by a Node.js & Express.js server. This handles the API requests securely and manages the business logic.
  • AI Integration: We utilized the @google/generative-ai SDK to integrate the Gemini 2.5 Flash Lite model, taking advantage of its fast inference times for near-instant email generation.
  • Database & ORM: We integrated Prisma ORM alongside PostgreSQL to establish a robust foundation for tracking template usage, user prompts, and generation history.

Challenges we ran into

  • Prompt Engineering: One of the biggest hurdles was tweaking the backend prompt construction to ensure the AI reliably respected the "Formality" slider and chosen tone without hallucinating extra details or ignoring the user's specific bullet points.
  • Full-Stack Wiring: Configuring the monorepo-style structure between the Vite frontend and the Express/Prisma backend—ensuring CORS, environment variables, and database connections all synced up perfectly—required careful debugging.
  • UI/UX Design: Designing an interface that feels highly premium and modern, yet remains simple and intuitive. Balancing the layout so the generated email and the input controls both have adequate space was a fun UI challenge.

Accomplishments that we're proud of

  • Blazing Fast Generation: Successfully integrating the Gemini 2.5 Flash Lite model to provide near-instantaneous, high-quality text generation.
  • Aesthetic Excellence: Creating a UI that genuinely wows the user. The combination of Tailwind CSS and Framer Motion makes the app feel alive, responsive, and highly professional.
  • Scalable Foundation: Implementing Prisma and PostgreSQL right out of the gate, which sets us up perfectly to add complex user data and analytics in the future without needing a massive rewrite.

What we learned

  • Working with LLMs: We gained a deeper understanding of how to structure prompts programmatically (e.g., dynamically injecting template strings and variables like tone and formality) to get consistent results from Gemini.
  • Prisma Integrations: We leveled up our database management skills by defining robust schemas and handling migrations with Prisma ORM.
  • Animation & State: We learned how to better manage complex React states alongside Framer Motion animations to ensure the UI updates smoothly when new AI data streams in.

What's next for Pro_Compose

  • Authentication & Accounts: Adding user login so individuals can save their generated drafts, view history, and store custom templates.
  • Contextual Replies: Allowing users to paste an email they received, so the AI can generate a context-aware reply rather than just drafting from scratch.
  • Browser Extension: Bringing Pro_Compose directly into Gmail, Outlook, or LinkedIn via a Chrome Extension so users can generate text without leaving their inbox.
  • Multi-Language Support: Extending the AI generation to automatically draft or translate professional emails for international communication.
Share this project:

Updates