DeliveryNotes+

Turn messy requirement notes into polished user stories—in seconds.

DeliveryNotes+ is an AI-powered SaaS that helps product owners and product managers turn raw business requirement notes into professional user stories, clear action items, and edge cases. Stop wrestling with format; focus on what matters.

---

What It Does

Paste your notes from a stakeholder call, email, or scratch doc. DeliveryNotes+ uses AI to:

• User stories — Structured as "As a <role>, I want to <goal> so that <benefit>" for both non-technical stakeholders and dev teams.
• Action items — Concrete next steps and follow-ups for every request.
• Edge cases — Top considerations and risks so nothing slips through.
• Estimates — Suggested time to complete based on complexity (on supported plans).
• Export — Download the result as a PDF for tickets, docs, or handoffs.

Plans (free, basic, premium) control output depth—e.g. acceptance criteria and risk lists on higher tiers.

---

Features

Feature	Description
Clerk auth	Sign in and plan-gated access (basic / premium).
Streaming output	AI response streams in real time via SSE.
PDF export	One-click export of the generated user story summary.
Plan-aware API	Backend tailors length and detail from the user’s subscription.

---

Tech Stack

• Frontend: Next.js (Pages Router), React, Tailwind CSS, Clerk, React Markdown, jsPDF
• Backend: Python, FastAPI, OpenAI, Clerk JWT validation
• Deploy: Single Docker image (Next.js static export + FastAPI on port 8000)

---

Quick Start

Run with Docker (recommended)

The app is designed to run as one service: FastAPI serves the static Next.js app and the /api/DeliveryNotes endpoint.

1. Set environment variables (see Environment variables below).

2. Build and run:

   docker build -t deliverynotes-plus .
   docker run -p 8000:8000 --env-file .env deliverynotes-plus

3. Open http://localhost:8000 in your browser.

Local development

• Frontend: From the project root, run npm install then npm run dev. Open http://localhost:3000.
  For the AI feature to work, the frontend must call the same API (e.g. by running the API locally and using a proxy or NEXT_PUBLIC_* API URL, depending on your setup).
• API: From the project root, run the FastAPI server (e.g. uvicorn api.server:app --reload --port 8000). Ensure .env is set so the API can reach Clerk and OpenAI.

---

Environment Variables

Variable	Purpose
NEXT_PUBLIC_CLERK_PUBLISHABLE_KEY	Clerk public key (frontend).
CLERK_SECRET_KEY	Clerk secret (backend auth).
CLERK_JWKS_URL	Clerk JWKS URL for JWT verification.
OPENAI_API_KEY	OpenAI API key for story generation.

Use these in .env for local runs and in your Docker/App Runner (or other) deployment config. Do not commit real secrets.

---

Project Structure

saas/
├── pages/           # Next.js pages (index, product)
├── api/
│   └── server.py    # FastAPI app: /api/DeliveryNotes, /health, static serve
├── Dockerfile       # Builds Next static export + Python image
├── .env             # Local env (do not commit secrets)
└── README.md

---

API

• POST /api/DeliveryNotes — Accepts JSON { product_owner, date_of_request, notes } and an Authorization: Bearer <Clerk JWT>. Streams the AI-generated user story summary as Server-Sent Events (SSE). Plan is read from the JWT to adjust model behavior.
• GET /health — Returns { "status": "healthy" } for load balancers and App Runner.

---

Next Steps

v1 — Deploy to AWS (ECR + App Runner)

Deploy the Docker image to AWS for a production-ready v1:

1. Push the image to Amazon ECR: Create an ECR repository, authenticate Docker to ECR, then build and push the deliverynotes-plus image.
2. Create an App Runner service: Point App Runner at your ECR image, set port 8000, and configure environment variables (Clerk keys, OpenAI key, etc.) in the service. Use the built-in health check on /health.
3. Access the app via the App Runner-assigned URL (or attach a custom domain in the next phase).

Immediate improvements (post–v1)

Improvement	Description
Custom domain	Add your own domain in App Runner settings.
Auto-deployment	Set up GitHub Actions to build, push to ECR, and update the App Runner service on push.
Monitoring	Add CloudWatch alarms for errors and latency.

Advanced enhancements (future versions)

Enhancement	Description
Database	Add Amazon RDS for data persistence (e.g. user story history, audit logs).
File storage	Use S3 for user uploads (e.g. attached requirement documents).
Caching	Use ElastiCache for performance (e.g. session or response caching).
CDN	Use CloudFront for global distribution and faster static assets.
Secrets Manager	Store sensitive data (API keys, DB credentials) in AWS Secrets Manager instead of env vars.

---

Learn More

• Next.js Docs
• Clerk
• FastAPI