How AI is used here, in plain words
The prospect you practice with is an AI, your card is written by an AI against a method written by people, and here is the rest in plain words.
Last updated October 24, 2026.
10before10™ is a software company. The prospect you practice with is an AI. On a practice call, the seller, buyer, owner or agent on the other end is a computer voice driven by an AI model. It is not a person. It is not a recording of a real person. No real seller or buyer is involved. We say this at the start of your first call and here.
Your card is written by an AI, against a method written by people. After each call, an AI model reads the transcript and writes your card: the one thing to work on, whether the call kept the ten rules of our method, and where you stand across ten skill areas. The founders wrote the method and the standards. The AI applies them; it did not invent them.
After a call, an AI may grade your lead sheet. If your organization uses the lead sheet, an AI compares what you wrote down with what the practice prospect said. It is practice feedback on listening. It never concerns a real person.
Forms training is written by an AI, from your organization's own form. If your brokerage, association or school uploads its forms, an AI reads each form and writes a walk-through: what each section says, what a licensee completes, how to explain it to a client, and common traps. It also writes quiz questions and plays a client who asks about a section. Your organization's manager reviews the walk-through before you see it and can change any part. It is training material, not legal advice.
Spoken answers are turned into text by an AI, and the audio is not kept. When you answer a client out loud, your recording goes to a speech-to-text AI. The text comes back and is graded. We keep the text. We do not keep the audio.
Assessments are graded by the same AI, and a person can always read the record. If your school or association turns on an assessment for a form, you answer a fixed set of client questions against a clock. An AI scores each answer on three things: accurate, plain, and staying on the licensee's side of the line (never giving legal advice). Your organization sets the pass mark. Your instructor can read every answer, every score and the feedback. Your organization, not 10before10, decides what a result means for its course.
What the AI is good at, and what it is not. It is consistent and patient. It is not infallible. A grade on a single call can be off by about one step either way. The AI can misjudge tone, miss context, or misquote you. If a card or a grade looks wrong, tell us at hello@10before10.com and a person will look at it. For an assessment, also tell your instructor.
Which AI services are involved. Anthropic (Claude) writes cards, walk-throughs, quiz and client questions, and grades answers. Google (Gemini) turns spoken answers into text and reads scanned form pages. ElevenLabs provides the voices: the practice prospect, the narrator and the client. The Privacy Policy lists each provider and what it receives.
Training on your data. We read practice transcripts, cards and the ratings you give us to improve the method, the practice prospect and the scoring. We do not use your calls, answers or cards to train any AI model that others use. Our AI providers process your data to give us the service.
What the AI never does. It never gives legal, tax or investment advice. It never decides anything about your license, your job or your pay. Any such decision is a person's decision under your organization's rules. It never contacts a real prospect for you.
Your choices. You can delete your history or your account. You can type instead of speaking. Questions: hello@10before10.com. A person answers.
This page is also published in Spanish. If the two versions differ, the English version governs.