Hello @bugun,
AI Engine doesn’t ship a model; pick OpenAI, Claude, etc. They all handle Turkish well, I believe! 😊
Here’s a super short playbook:
If you have ≤ 20 products, paste core product info + links right into the chatbot Instructions:
Tüm yanıtlarda Türkçe kullan. Sadece aşağıdaki ürün bilgilerinden yararlan:
1) Ürün: X100 Kulaklık Fiyat: 1.299 ₺ Özellikler: ANC, BT 5.3, 40 saat Link: https://site.com/urun/x100
2) Ürün, etc...
Keep each product to 4–6 lines to avoid token bloat.
If you have many products, you’ll need to use Knowledge (Vector Database) to sync your WooCommerce products. It’s more complex, but it’s the approach most sites and systems use. You can learn more here.
Alternatively, if you’re a developer (or have one around), you can use Function Calling. Create a function via Code Engine that retrieves all your products or specific product info, then connect your chatbot to those functions. You can learn more about it here.
My advice: take your time 😌 Start by playing with the instructions and experimenting with just a few products first. Jumping to a Vector Database is quite a big step, and you’ll gradually discover how larger sites manage their knowledge base and workflows. It’s not a one-click kind of magic, but that’s part of the fun.
