Build a Developer Roadmap Generator with OpenAI and React

Build a Developer Roadmap Generator with OpenAI and React

In the fast-paced world of software development, the path to learning a new stack can feel overwhelming. With a vast array of programming languages, frameworks, and specialties, it's tough to know where to start and what skills to prioritize.

Now imagine if you had access to a personalized plan, tailored to your interests and goals?

In this tutorial, we will show you how to build a Developer Roadmap Generator with OpenAI and React using Platformatic AI Warp.

Platformatic AI Warp is a gateway that simplifies integrating AI providers, including OpenAI, into your application. This guide will teach you how to set up a Platformatic AI Warp application, add GitHub OAuth2 for authentication, create a React frontend to interact with AI Warp, and deploy your application to

The complete code is available on GitHub, and a live version of this application can be found here.


To follow this tutorial, you will need:

  • Basic understanding of Platformatic and JavaScript

  • Node.js LTS version

  • Docker installed on your machine for easy deployment

A GitHub account is required for OAuth2 authentication

Create an AI Warp Application

To create and run your AI Warp application, first navigate to the GitHub for AI Warp and run the application using the steps below:

npx create-platformatic@latest
  • Select Application, then @platformatic/ai-warp

  • Enter your ai-app name, in our case, developer-roadmap-generator

  • Select your AI provider.

  • Enter the model you want to use

  • Enter the API key if you are using an online provider (OpenAI).

Run the command to start AI Warp:

npm start

You will see your AI warp application when you open http://localhost:3042/ in your browser.

NOTE: Check out our guide to running a local model of Llama2 and add it as an AI provider to your platformatic.json file.

Setting Up Prompt Decorators

With Platformatic AI Warp, you can pre-set a system prompt to have more control over how an LLM service is used by adding promptDecorators.

To do this, navigate to the platformatic.json file in your ai-warp application's services/ai folder and update it as shown below.

// services/ai/platformatic.json 

"promptDecorators": {
      "prefix": "You are an expert career advisor specializing in creating detailed and structured developer roadmaps. You will generate an in-depth and comprehensive roadmap for users in markdown format including stages, skills, and resources required to progress from a beginner to an advanced developer.\nThe specific roadmap you need to create is: ",
      "suffix": "Please ensure the roadmap is very detailed, covering key areas such as foundational knowledge, programming languages, frameworks, tools, best practices, and continuous learning strategies."  

Setting up Rate Limiting

We need to set up rate limits and disable AI Warp's default homepage. To do this, navigate to platformatic.json inside the services/ai folder and add the following rate limits.

// services/ai/platformatic.json 

  "showAiWarpHomepage": false,
  "rateLimiting": {
    "max": 100,
    "timeWindow": "1 minute",
    "maxByClaims": [
        "claim": "userType",
        "claimValue": "premium",
        "max": 1000

Creating a Client for AI Warp OpenAPI Server

To create a frontend client for a remote OpenAPI server for AI Warp, run the command below in the services/ai directory:

npx platformatic client --frontend --name AI

This will create a folder in your project directory with three files:

  • AI-types.d.ts: Contains types and interfaces for your AI Warp

  • AI.mjs: the client, a JavaScript implementation of your OpenAPI server, including routes and endpoints from AI Warp.

  • AI.openapi.json: A JSON object of your client, including AI Warp endpoints.

NOTE: Make sure to change the URL and directory name in the command to your application URL and name.

Adding Authentication with GitHub OAuth2

To ensure that only authenticated users can generate developer roadmaps with AI, we will integrate GitHub OAuth2. This section will guide you through the process of setting up and configuring GitHub OAuth2 in your application.

First, follow the steps outlined in the GitHub documentation to create a new OAuth2 application and save your GitHub OAuth credentials in your .env file.

// .env


Install Fastify Packages

Install fastify/oauth2, fastify-cookie, fastify-pluginpackages in the root of our application with the command.

npm i @fastify/oauth2 fastify-cookie fastify-plugin undici

Follow the steps outlined in the GitHub documentation and create a new OAuth2 application.

In the services/ai folder, create a new folder auth, and inside it, create a file utils.js and add the code below:

// services/ai/auth/utils.js

const { request } = require("undici");

async function callGHEndpoint({ path, method, body, accessToken }) {
  const res = await request(`${path}`, {
    method: method.toUpperCase(),
    headers: {
      authorization: `Bearer ${accessToken}`,
      accept: "application/json",
      "X-GitHub-Api-Version": "2022-11-28",
      "User-Agent": "A Platformatic App",
  return await res.body.json();
module.exports = {

Here, we create a utility function called callGHEndpoint, which authenticates requests to the GitHub API. We construct the request URL, set the request headers, and parse the response as JSON.

Next, we integrate the OAuth2 plugin into our AI Warp application. Inside the auth folder, create a new file, github.js, and add the code blocks below.

// services/ai/auth/github.js

"use strict";
/// <reference path="../global.d.ts" />

const { callGHEndpoint } = require("./utils");
const oauthPlugin = require("@fastify/oauth2");

const CALLBACK_URL = process.env.CALLBACK_URL || "http://localhost:3042";

/** @param {import('fastify').FastifyInstance} app */
module.exports = async function (app, opts) {
  app.register(oauthPlugin, {
    name: "githubOAuth2",
    credentials: {
      client: {
        id: opts.GITHUB_OAUTH_CLIENT_ID,
        secret: opts.GITHUB_OAUTH_CLIENT_SECRET,
      auth: oauthPlugin.GITHUB_CONFIGURATION,
    startRedirectPath: "/login/github",
    callbackUri: `${CALLBACK_URL}/login/github/callback`,
    cookie: {
      path: "/",
      secure: true,
      sameSite: "none",

  app.get("/login/github/callback", async (req, res) => {
    try {
      const { token } =
        await app.githubOAuth2.getAccessTokenFromAuthorizationCodeFlow(req);
      const githubUser = await callGHEndpoint({
        path: "user",
        accessToken: token.access_token,
        method: "GET",
        body: {},
      const user = {
        username: githubUser.login,
        image: githubUser.avatar_url,

      const userToken = Buffer.from("base64");

      const redirectUrl = `${CALLBACK_URL}?username=${encodeURIComponent(user.username)}&token=${userToken}`;
    } catch (error) {
      res.send("Authentication failed");

Here, we added GitHub OAuth2 authentication to our AI Warp application to enable users to log in with GitHub. First, we imported the utility function to call the GitHub API. Next, we registered the @fastify/oauth2 with a Fastify instance and configured it with our GitHub credentials. The callback URI will handle the response from GitHub authenticating and set a cookie for the application OAuth2 flow.

In the second code block, we defined a route handler for /login/github/callback to handle the OAuth2 callback, which gets an access token from a user after authorizing the application. It then fetches the user’s profile information and creates a user object. Finally, we redirect the user to the CALLBACK_URL, our AI application's homepage.

With this approach, our application initiates the OAuth2 flow when users attempt to log in, redirects them to GitHub for authentication, and handles the callback to complete the login process. This way, only authenticated users can use the developer roadmap application.

Creating Plugins for Custom OAuth2 Flow Handling

In this section we will add a custom method to handle actions after the OAuth2 flow is completed. We will use the fastify-plugin module to create a Fastify plugin that manages user login and redirection after successful authentication.

Create an Authentication Plugin

First, in the services/ai folder, create a new file named authentication.js and add the following code.

// services/ai/plugins/authentication.js

const fp = require("fastify-plugin");

async function plugin(app, options) {
  app.decorate("afterOAuth2Flow", async (user, externalId, req, res) => {
    req("user", user);
    const loggedInUser = await app.loginUser({
      username: user.username,
    });{ loggedInUser });
    if (app.config.DEV === true) {
      return res.redirect(app.config.PLT_MAIN_URL);
    return res.redirect(app.config.PLT_MAIN_URL);

plugin[Symbol.for("skip-override")] = true;

module.exports = fp(plugin, {
  name: "authentication",

We extended our Fastify application by adding a custom OAuth2 flow-handling plugin to manage user login and redirection after authentication. We defined a plugin function that adds a custom after OAuth2Flow method to the Fastify instance.

The after OAuth2Flow method handles the authenticated user's details, attaches the user object to the request, and logs the user in using their email, username, and external ID. It then logs the logged-in user information and redirects the user to the main URL, the application homepage.

Finally, we exported the plugin using fastify-plugin, naming it 'authentication' to integrate seamlessly with our AI application.

Setting up Environment Variables

Before building our React application, let’s inspect and setup our environment variables, navigate to the .env file in the root folder of your application and verify that it looks similar to the one below:


Building a React application

Follow the official Vite documentation to create a React application with the latest version of React. Run this command in the services/ai folder of your application:

npm create vite@latest

Here, you will be presented with a few options — for the sake of this tutorial:

  • Project name: … Ai-frontend

  • Package name: … ai-frontend

  • Select a framework: › React

  • Select a variant: › JavaScript

Once you have created the project, go to the project directory and install the packages:

npm i @heroicons/react axios next-themes react-hot-toast react-markdown react-modal remark-gfm

Creating Utility Functions for the React App

Create a new file utils.js, inside a utils folder in the src folder of your frontend application and add the code block below:

// // services/ai/src/utils/utils.js 

import { useState, useCallback } from "react";
import { toast } from "react-hot-toast";
import { prompt as apiPrompt, setBaseUrl } from "../../AI.mjs";

We imported the postApiV1Prompt function and the setBaseUrl function from our API utilities module at ../../AI.mjs. Using setBaseUrl, we configure our application to use the base URL specified in our environment variables (import.meta.env.VITE_AI_WARP_URL).

This approach allows us to dynamically set the base URL for API requests, which is particularly useful for handling different environments such as development, testing, and production.

You can set the base URL directly in your application code using setBaseUrl, either by hardcoding or fetching it from an environment configuration file (.env).

Fetching a Prompt from AI Warp

Let’s create a utility function to fetch a prompt:

// services/ai/src/utils/utils.js

export const useFetchPrompt = () => {
  const [prompt, setPrompt] = useState("");
  const [response, setResponse] = useState("");
  const [loading, setLoading] = useState(false);

  const fetchPrompt = useCallback(async (userPrompt) => {

    try {
      const fetchedPrompt = await apiPrompt({ prompt: userPrompt });
      const responses = [...response, fetchedPrompt.response];
      toast.success("Prompt fetched successfully!");
    } catch (error) {
      console.error("Failed to load prompt:", error);
      toast.error("Failed to fetch data.");
    } finally {
  }, []);

  return {

Here, we created the useFetchPrompt hook function to store our application's prompt, response, and loading status for our AI application.

Handling Prompt Submit

In this section, we will create the function for submitting our prompt on the front end.

// services/ai/src/utils/utils.js

export const handlePromptSubmit = async (prompt, setResponse, setLoading) => {
  try {
    const res = await fetch(`${import.meta.env.VITE_AI_URL}/api/v1/stream`, {
      method: "POST",
      headers: {
        "Content-Type": "application/json",
      body: JSON.stringify({ prompt }),

    if (!res.ok) {
      const errorResponse = await res.json();
      setResponse(`Error: ${errorResponse.message} (${res.status})`);
    const reader = res.body.getReader();
    const decoder = new TextDecoder();
    let completeResponse = "";

    let loading = true;
    while (loading) {
      const { done, value } = await;
      if (done) break;

      const decodedValue = decoder.decode(value, { stream: true });
      const lines = decodedValue.split("\n");
      for (let i = 0; i < lines.length; i++) {
        const line = lines[i];
        if (line.startsWith("event: ")) {
          const eventType = line.substring("event: ".length);
          const dataLine = lines[++i];
          if (dataLine.startsWith("data: ")) {
            const data = dataLine.substring("data: ".length);
            const json = JSON.parse(data);
            if (eventType === "content") {
              completeResponse += json.response;
            } else if (eventType === "error") {
              setResponse(`Error: ${json.message} (${json.code})`);

  } catch (error) {
    console.error("Failed to fetch data:", error);
    setResponse("Failed to fetch data.");
  } finally {

export const handleKeyDown = (event, handlePromptSubmitCallback) => {
  if (event.key === "Enter" && !event.shiftKey) {

This function sends a user prompt to the /API/v1/stream of the base URL and handles the response. It reads and processes the data as streams, parses, and appends the results for a request.

We then log any errors during the request and send them to our frontend as a response. Any errors or exceptions during the request or processing phases are caught and logged, and a generic error message is displayed to the user.

The handleKeyDown function checks if the Enter key is pressed without the Shift key. If so, it prevents the default action (e.g., form submission) and triggers the handlePromptSubmitCallback function used to submit data.

Creating a Terms Component

In the src folder, create a new file terms.jsx, where you will add a terms and conditions modal for our application.

// services/ai/src/components/terms.jsx

import { useState } from "react";
import Modal from "react-modal";
import "../App.css";

const TermsOfService = () => {
  const [modalIsOpen, setModalIsOpen] = useState(false);
  const currentDate = new Date().toLocaleString();

  const openModal = () => setModalIsOpen(true);
  const closeModal = () => setModalIsOpen(false);

  return (
      <footer className="fixed bottom-0 w-full bg-gray-800 text-white py-2 text-center">
        <button onClick={openModal} className="text-sm">
          Terms of Service
        contentLabel="Terms of Service"
        className="modal bg-white p-6 rounded-lg shadow-lg max-w-2xl mx-auto my-10"
        overlayClassName="overlay fixed top-0 left-0 right-0 bottom-0 bg-black bg-opacity-50 flex items-center justify-center"
        <div className="max-w-2xl mx-auto">
          <h1 className="text-3xl font-bold mb-6">Terms of Service</h1>
          <p className="mb-4">
            Welcome to Platformatic AI roadmap generator. By using our service,
            you agree to the following terms:
          <h2 className="text-2xl font-semibold mb-4">
            1. Acceptance of Terms
          <p className="mb-4">
            By accessing and using our services, you accept and agree to be
            bound by the terms and provisions of this agreement.
          <h2 className="text-2xl font-semibold mb-4">
            2. Description of Service
          <p className="mb-4">
            Our service provides users with access to resources and tools for
            managing and interacting with AI models.
          <h2 className="text-2xl font-semibold mb-4">3. User Obligations</h2>
          <p className="mb-4">
            You must provide accurate registration information and keep it up to
            date. You are responsible for maintaining the confidentiality of
            your account.
          <h2 className="text-2xl font-semibold mb-4">4. Termination</h2>
          <p className="mb-4">
            We reserve the right to terminate or suspend your account at our
            sole discretion for conduct that we believe violates these terms.
          <p className="mt-8">Last updated: {currentDate}</p>
            className="mt-4 bg-gray-800 text-white py-2 px-4 rounded-lg"

export default TermsOfService;

Building our React Application

Update the App.jsx with the utility functions and add a React component for our application.

// services/ai/src/App.jsx

import { useEffect, useState } from "react";
import "./App.css";
import { useFetchPrompt, handlePromptSubmit } from "./utils/utils.js";
import ReactMarkdown from "react-markdown";
import remarkGfm from "remark-gfm";

function App() {
  const {
  } = useFetchPrompt();
  const [isAuthenticated, setIsAuthenticated] = useState(false);
  const [user, setUser] = useState(null);

  useEffect(() => {
  }, [fetchPrompt]);

  useEffect(() => {
    const query = new URLSearchParams(;
    const token = query.get("token");
    const username = query.get("username");
    const fullname = query.get("fullname");
    const image = query.get("image");
    const email = query.get("email");

    if (token) {
      setUser({ username, fullname, image, email });
  }, []);

  const handlePromptChange = (event) => {

Here, we import our utility functions, useFetchPrompt, and handlePromptSubmit to handle API requests. ReactMarkdown and remarkGfm will render the response from our API as markdown with GitHub-flavored markdown.

We then restructured the useFetchPrompt custom hook. The isAuthenticated state checks if the user is authenticated, and the user state stores the authenticated user details — the handlePromptChange updates the prompt state whenever the input field value changes.

// services/ai/src/App.jsx

const submit = () => handlePromptSubmit(prompt, setResponse, setLoading);

  const handleLogin = () => {
    window.location.href = `${import.meta.env.VITE_AI_URL}/login/github`;

  return (
    <div className="min-h-screen bg-gray-900 text-white flex flex-col items-center justify-center">
      <div className="p-6 bg-gray-800 rounded-lg shadow-lg w-full max-w-lg">
        {!isAuthenticated ? (
          <div className="mb-4">
            <h1 className="text-2xl font-bold mb-2">
              Platformatic AI Roadmap Generator
              className="w-full bg-indigo-600 text-white py-2 px-4 rounded-lg hover:bg-indigo-700 transition duration-200 mb-4"
              Login with GitHub
        ) : (
            <div className="mb-4">
                className="block text-xl font-bold mb-2"
                Platformatic Roadmap Generator
            <div className="mb-4">
                className="w-full p-2 text-black rounded-lg focus:outline-none focus:ring-2 focus:ring-indigo-500"
                placeholder="Get your AI roadmap here..."
            <div className="mb-4">
                className="w-full bg-indigo-600 text-white py-2 px-4 rounded-lg hover:bg-indigo-700 transition duration-200"
                {loading ? "Generating..." : "Get Roadmap"}
            {!loading && response && (
              <div className="p-4 bg-gray-700 rounded-lg mt-4 w-full max-h-96 overflow-y-auto">
                <div id="messages" className="prose prose-invert">
                  <ReactMarkdown remarkPlugins={[remarkGfm]}>

export default App;

Here, we authenticate users and redirect them to the developer roadmap homepage. A textarea input field allows the authenticated user to enter a prompt and receive an AI-personalized roadmap. The submit button triggers the handlePromptSubmit function to send the prompt to our AI provider API. We then render the response as a markdown using the ReactMarkdown with remarkGfm.

Restructuring Static Files

In this section, we will restructure our application and serve our frontend application as static files using @fastify/static.

First, move your frontend application's src and public folders into the services/ai folder and install all your frontend packages in the package.json file in the services/ai folder.

Rendering the Static Files

To serve the frontend application as a static file, first install the @fastify/static package.

npm i @fastify/static

Create a new file named static.js in the plugins folder in your services/ai project directory and add the code below:

const fp = require("fastify-plugin");
const path = require("path");

async function plugin(app, options) {
  app.register(require("@fastify/static"), {
    root: path.join(__dirname, "..", "dist"),
    prefix: "/",

module.exports = fp(plugin, {
  name: "static",

Here, we created a plugin function that registers the @fastify/static plugin with a Fastify instance. We then set the root directory for static files to the dist folder and serve the files from the root URL path (/).

Dockerizing and Running your Application

Before we deploy our application to, we first need to dockerize it. To do this, create a Dockerfile in the root directory of your project and add the script.

// Dockerfile

FROM node:20-alpine
# RUN corepack enable && corepack prepare pnpm@latest --activate
ENV APP_HOME=/home/app/node/
COPY package.json package.json
COPY package-lock.json package-lock.json
COPY platformatic.json platformatic.json
COPY services services
RUN npm install
RUN cd services/ai && \
  npm install && \
  npm run build
CMD ["npm", "start"]

In this Dockerfile, we use the node:20-alpine base image for our application. We added a build time argument VITE_AI_URL which is an environment variable inside the container. We then set the working directory to /home/app/node/, where application files are copied.

npm install installs the dependencies in our package.json . We then exposed port 3042 for running the application. The CMD ["npm", "start"] command launches the application when the container starts.

To build the Docker image, use the command:

docker build -t roadmap-generator:latest .

Then, run your entire application with the command:

docker run --env-file .env -p 3042:3042 roadmap-generator:latest

You can access the application at http://localhost:3042

Click on the Login with GitHub button and you will be redirected back to the homepage.

Deploying to

Now we can deploy our application to To do so, first, follow the Fly documentation on installing Fly on your local machine.

Run the command to deploy your application.

fly deploy

Follow the steps outlined in the Fly documentation to complete deployment.

Wrapping Up

In this tutorial, we built a developer roadmap generator using Platformatic AI Warp, integrated GitHub OAuth2 authentication and deployed the application with Docker and We also learned to restructure our application to serve static files. By following these steps, your developer roadmap application is now deployed on

To improve the application, you can extend it by adding PDF downloads for personalized roadmaps.

Further Resources