Add TODOs

.github/workflows/build.yaml

@@ -0,0 +1,48 @@
+name: Build Binaries
+
+on: workflow_dispatch
+
+env:
+  CARGO_TERM_COLOR: always
+
+jobs:
+  linux:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@v3
+      - name: Build for Linux
+        run: cargo build --release
+      - name: Upload artifact
+        uses: actions/upload-artifact@v4
+        with:
+          name: tera-linux-x86-64
+          path: target/release/tera
+
+  windows:
+    runs-on: windows-latest
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@v3
+      - name: Build for Windows
+        run: cargo build --release
+      - name: Upload artifact
+        uses: actions/upload-artifact@v4
+        with:
+          name: tera-windows-x86-64
+          path: target/release/tera
+
+  mac:
+    runs-on: macos-latest
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@v3
+      - name: Install compilation target
+        run: rustup target add aarch64-apple-darwin
+      - name: Build for Mac ARM
+        run: cargo build --release --target aarch64-apple-darwin
+      - name: Upload artifact
+        uses: actions/upload-artifact@v4
+        with:
+          name: tera-mac-arm64
+          path: target/aarch64-apple-darwin/release/tera

README.md

@@ -1,6 +1,8 @@
 # Tera
 Tera is an AI assistant which is tailored just for you and runs fully locally.
 
+> I just published a [Medium Article](https://medium.com/@isalapiyarisi/lets-build-a-standalone-chatbot-with-phi-2-and-rust-48c0f714f915) about how I built Tera. Check it out if you're interested.
+
 ## Build
 
 1. Make sure you have all the dependencies installed.
@@ -50,5 +52,11 @@ Options:
 4. **Motivational Coach**: Feed Tera with motivational quotes and ask it to inspire you when you need a boost.
 4. **Home Inventory Manager**: Tell Tera about your home inventory and ask it to remind you when you're running low on supplies.
 
+### TODOs
+- [ ] Make use of SurrealDB's [vector indexing](https://www.youtube.com/watch?v=2MmyE_iohEs) to improve performance once it's available.
+- [ ] Publish prebuilt binaries.
+- [ ] Add CUDA and Metal support for faster inference.
+- [ ] Remove ffmpeg dependency.
+
 ## Licence
 [AGPL-3.0-or-later](LICENSE)