JevMade Sign in
← Back to videos

JevMade field notes / Video guide

JEV: É Aqui Que Eu Vejo Dinheiro de Verdade com IA

Kelvin Cleto describes adding Jev to a sales copilot that follows meeting transcripts. He uses small questions about answers, objections and the sales script to decide when the application needs an LLM's written suggestion.

Original by Kelvin CletoAgent workflowsIntermediate23 min 32 sec Published

Before you press play

What you’ll find in the video

  1. Keep speech recognition separate, then use the organized transcript as input for the decision step.
  2. Ask whether an answer is complete or a prepared question is still suitable before requesting a written suggestion from an LLM.
  3. Treat the projected reduction from R$20–30 to R$5 per meeting as a goal for the next pilot, not a measured saving.
Worth knowing

Kelvin says he began the integration that day and planned a one-salesperson pilot for the next day. His prices, latency figures and cost calculation are informal reports; automatic captions make some numeric units unclear. The 0.80 cutoff is an example, not a validated policy. The video promotes his own business and workshop. Customer data handling and sales outcomes were not verified.

JEV: É Aqui Que Eu Vejo Dinheiro de Verdade com IA

Keep this for later

Sign in to bookmark experiments, guides and videos, and keep notes only you can see.

Continue to sign in

We’ll bring you back to this listing.