👂 Ears — every voice note, understood
Transcribes every voice note with Whisper (large-v3-turbo) and replies to the content — not with an ‘I can't listen to audio.’ Works on WhatsApp, Telegram, Instagram, and Messenger.
A 40-second voice note, a photo of the receipt, a screenshot of the product — if your bot can't understand them, the customer leaves. This one does.
› included with Forja + · in your cloud, with your data
// same engine across every channel
The bot listens and sees within the same conversation — it never tells anyone to ‘just type out your question instead.’
Transcribes every voice note with Whisper (large-v3-turbo) and replies to the content — not with an ‘I can't listen to audio.’ Works on WhatsApp, Telegram, Instagram, and Messenger.
With Forja +, the bot sees images with that same brain: a photo of a product (‘do you have any of these?’), a payment receipt, a location. And it replies to what it actually sees, not blindly.
The customer explains their order in a 40-second voice note and sends the receipt as a photo. The bot understands both and confirms — without ever asking them to ‘type it out.’
// Whisper transcribed the audio; its own brain (Forja +) saw the receipt.