How We Handle Your Data

Last updated: February 2026

Role Play For Impact helps you practice fundraising conversations with an AI donor. To do that well, we need to collect and store your practice sessions. This policy explains exactly what we collect, who can see it, and what control you have over it. We've written this in plain English because you shouldn't need a lawyer to understand how your data is handled.

Your practice sessions belong to you. We store them so you can review and learn from them. If you're working with a coach, they can see your sessions too — that's how coaching works. We also use anonymized, aggregated data to make the product better for everyone.

What We Collect

Account information

Your name, email, and profile picture from Google sign-in. We use this to identify you and personalize your experience.

Practice sessions

When you practice, we record:

  • The full conversation transcript (what you said, what the AI said)
  • Session details: duration, scenario name, donor persona, organization context
  • Automatic coaching analysis: scores, strengths, and growth areas generated by AI after each session

Scenarios you create

Any custom donor personas, organizations, or scenarios you build are saved to your account.

Usage data

Basic analytics: when you log in, which features you use, how often you practice. This helps us understand how people use the product and where to improve it.

What we don't collect

We don't store audio recordings — voice processing happens in real-time and only the text transcript is saved. We don't collect payment information directly — that's handled by our payment processor.

Who Can See Your Data

You

You always have full access to your practice sessions, transcripts, analysis, and scenarios. You can view, export, or delete any of your data.

Your coach (if assigned)

If you're working with a coach, they can see all your practice sessions, transcripts, and coaching analysis. This is how coaching works — your coach needs to see your practice to help you improve. If you're not assigned to a coach, no one else sees your sessions.

Our team

For Impact team members may access your data for:

  • Technical support (helping you troubleshoot issues)
  • Product improvement (understanding how people use the product)
  • Quality assurance (ensuring the AI is working correctly)

We access individual sessions only when necessary. We don't read your transcripts for fun.

Third parties

We use these services to run the product:

  • Vapi — processes voice calls in real-time
  • Supabase — stores your data
  • Anthropic (Claude) — generates AI responses and coaching analysis

These providers process your data according to their own privacy policies. We've chosen providers with strong privacy practices.

How We Improve the Product Using Your Data

Automatic coaching analysis

After each practice session, we automatically analyze your conversation using AI. This generates scores, identifies strengths, and suggests areas for improvement. This analysis is stored with your session and visible to you (and your coach, if assigned).

Anonymized product improvement

We use aggregated, anonymized data from all users to make the product better:

  • Better AI donors — Understanding common conversation patterns helps us make the AI more realistic
  • Better coaching feedback — Seeing what good practice looks like helps us improve the automated analysis
  • Benchmarks — Aggregate performance data helps us define what “good” looks like so we can give you meaningful scores

What “anonymized” means

Before using data for product improvement, we strip out identifying information: your name, your organization's name, specific donor names mentioned, and other personal details. The goal is to learn patterns, not to study you specifically.

What's coming

We're building toward stronger data isolation:

  • Fully anonymized training pipelines (in development)
  • Per-organization data firewalls for enterprise accounts (planned)

We'll update this policy as these features ship.

Your Data, Your Controls

View and export

You can view any of your practice sessions, transcripts, and coaching analysis at any time. You can export transcripts as text or JSON files from your session history.

Delete sessions

You can delete individual practice sessions from your history. Deletion removes the transcript, analysis, and all associated metadata. This is permanent.

Delete your account

You can delete your entire account from settings. This removes:

  • Your profile information
  • All practice sessions and transcripts
  • All saved scenarios
  • All coaching analysis

If you were assigned to a coach, your sessions will no longer appear in their dashboard.

What deletion doesn't cover

If your data was already anonymized and included in aggregate product metrics before you deleted it, that anonymized data remains. Once information is stripped of identifying details and mixed into aggregate patterns, it's no longer tied to you and can't be individually removed.

Data portability

You can export your data at any time. We don't lock you in.

How We Protect Your Data

Infrastructure

Your data is stored in Supabase, which provides encryption at rest and in transit. Our application is served over HTTPS. Authentication is handled through Google OAuth — we never see or store your Google password.

Access controls

The application enforces role-based access. Users see only their own data. Coaches see only their assigned clients' data. Admins have broader access for support and operations.

What we don't do

We don't sell your data. We don't share it with advertisers. We don't use your identifiable data for any purpose other than providing and improving this product.

We take special care with prospect and funder information. Donor names, giving histories, and relationship details that come up in your practice sessions are never cross-referenced, mined, or aggregated across users. Your prospect intelligence stays yours. We understand that fundraiser-donor relationships are built on trust, and we will never do anything that could compromise that trust or create even the appearance that funder data is being shared or exploited.

Where We're Headed

This policy reflects where we are today. We're building toward stronger protections:

Anonymized training pipelines

Currently, product improvement happens through manual review of aggregated patterns. We're building automated pipelines that strip identifying information before any data is used for product improvement — removing names, organizations, and contextual details that could identify you or your prospects.

Per-organization data firewalls

For enterprise accounts, we're building the ability to fully isolate your organization's data from other users. Your sessions, scenarios, and practice data would exist in a logically separate environment.

Self-service data controls

We're adding the ability to delete individual sessions and your full account directly from settings, without needing to contact support.

We'll update this policy as each of these ships. If a change meaningfully affects how your data is handled, we'll notify you.