diff --git a/.github/workflows/ci.yml b/.github/workflows/ci.yml
new file mode 100644
index 0000000..f4abd1d
--- /dev/null
+++ b/.github/workflows/ci.yml
@@ -0,0 +1,29 @@
+name: CI
+
+on:
+  push:
+    branches: [main]
+  pull_request:
+
+jobs:
+  build:
+    name: Type-check and build
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v4
+
+      - name: Setup Node
+        uses: actions/setup-node@v4
+        with:
+          node-version: 20
+          cache: npm
+
+      - name: Install dependencies
+        run: npm ci
+
+      - name: Type-check
+        run: npm run check
+
+      - name: Build
+        run: npm run build
diff --git a/client/src/components/shared/LiveAudioDiscernment.tsx b/client/src/components/shared/LiveAudioDiscernment.tsx
index 4f4e664..84edba4 100644
--- a/client/src/components/shared/LiveAudioDiscernment.tsx
+++ b/client/src/components/shared/LiveAudioDiscernment.tsx
@@ -356,7 +356,7 @@ export function LiveAudioDiscernment() {
 
   const mediaRecorderRef = useRef<MediaRecorder | null>(null);
   const chunksRef = useRef<Blob[]>([]);
-  const streamIntervalRef = useRef<NodeJS.Timer | null>(null);
+  const streamIntervalRef = useRef<ReturnType<typeof setInterval> | null>(null);
 
   const handleConsent = () => {
     localStorage.setItem(CONSENT_STORAGE_KEY, "true");
diff --git a/package.json b/package.json
index 6645069..50de9bb 100644
--- a/package.json
+++ b/package.json
@@ -2,7 +2,7 @@
   "name": "rest-express",
   "version": "1.0.0",
   "type": "module",
-  "license": "MIT",
+  "license": "BUSL-1.1",
   "scripts": {
     "dev:client": "vite dev --port 5000",
     "dev": "NODE_ENV=development tsx server/index.ts",
diff --git a/server/replit_integrations/audio/index.ts b/server/replit_integrations/audio/index.ts
index 8d2a257..0999dd8 100644
--- a/server/replit_integrations/audio/index.ts
+++ b/server/replit_integrations/audio/index.ts
@@ -1,4 +1,3 @@
-export { registerAudioRoutes } from "./routes";
 export {
   openai,
   detectAudioFormat,
diff --git a/server/replit_integrations/audio/routes.ts b/server/replit_integrations/audio/routes.ts
deleted file mode 100644
index 6feaff0..0000000
--- a/server/replit_integrations/audio/routes.ts
+++ /dev/null
@@ -1,85 +0,0 @@
-import express, { type Express, type Request, type Response } from "express";
-import { chatStorage } from "../chat/storage";
-import { openai, speechToText, ensureCompatibleFormat } from "./client";
-
-// Body parser with 50MB limit for audio payloads
-const audioBodyParser = express.json({ limit: "50mb" });
-
-export function registerAudioRoutes(app: Express): void {
-  // Send voice message and get streaming audio response
-  // Auto-detects audio format and converts WebM/MP4/OGG to WAV
-  // Uses gpt-4o-mini-transcribe for STT, gpt-audio for voice response
-  app.post("/api/voice-conversations/:id/messages", audioBodyParser, async (req: Request, res: Response) => {
-    try {
-      const conversationId = parseInt(req.params.id as string);
-      const { audio, voice = "alloy" } = req.body;
-
-      if (!audio) {
-        return res.status(400).json({ error: "Audio data (base64) is required" });
-      }
-
-      // 1. Auto-detect format and convert to OpenAI-compatible format
-      const rawBuffer = Buffer.from(audio, "base64");
-      const { buffer: audioBuffer, format: inputFormat } = await ensureCompatibleFormat(rawBuffer);
-
-      // 2. Transcribe user audio
-      const userTranscript = await speechToText(audioBuffer, inputFormat);
-
-      // 3. Save user message
-      await chatStorage.createMessage(conversationId, "user", userTranscript);
-
-      // 4. Get conversation history
-      const existingMessages = await chatStorage.getMessagesByConversation(conversationId);
-      const chatHistory = existingMessages.map((m) => ({
-        role: m.role as "user" | "assistant",
-        content: m.content,
-      }));
-
-      // 5. Set up SSE
-      res.setHeader("Content-Type", "text/event-stream");
-      res.setHeader("Cache-Control", "no-cache");
-      res.setHeader("Connection", "keep-alive");
-
-      res.write(`data: ${JSON.stringify({ type: "user_transcript", data: userTranscript })}\n\n`);
-
-      // 6. Stream audio response from gpt-audio
-      const stream = await openai.chat.completions.create({
-        model: "gpt-audio",
-        modalities: ["text", "audio"],
-        audio: { voice, format: "pcm16" },
-        messages: chatHistory,
-        stream: true,
-      });
-
-      let assistantTranscript = "";
-
-      for await (const chunk of stream) {
-        const delta = chunk.choices?.[0]?.delta as any;
-        if (!delta) continue;
-
-        if (delta?.audio?.transcript) {
-          assistantTranscript += delta.audio.transcript;
-          res.write(`data: ${JSON.stringify({ type: "transcript", data: delta.audio.transcript })}\n\n`);
-        }
-
-        if (delta?.audio?.data) {
-          res.write(`data: ${JSON.stringify({ type: "audio", data: delta.audio.data })}\n\n`);
-        }
-      }
-
-      // 7. Save assistant message
-      await chatStorage.createMessage(conversationId, "assistant", assistantTranscript);
-
-      res.write(`data: ${JSON.stringify({ type: "done", transcript: assistantTranscript })}\n\n`);
-      res.end();
-    } catch (error) {
-      console.error("Error processing voice message:", error);
-      if (res.headersSent) {
-        res.write(`data: ${JSON.stringify({ type: "error", error: "Failed to process voice message" })}\n\n`);
-        res.end();
-      } else {
-        res.status(500).json({ error: "Failed to process voice message" });
-      }
-    }
-  });
-}