Our principles
- Your data stays yours. We use client data only to deliver the agreed work, and never to train models for anyone else.
- Least access. We ask only for the access a task needs, and remove it when the work ends.
- People stay accountable. AI systems we build include human review where decisions need judgment.
- No surprises. Data handling, hosting, and security requirements are agreed with you in writing before work begins.
Data handling in projects
- Data processing terms, including a data processing agreement where required, and NDAs, are agreed before we receive any client data.
- We prefer working in your environment or on representative sample data. Where we need copies, we keep them in access-controlled storage and delete them when the engagement ends, or as agreed.
- Personal and sensitive data is minimized, masked, or excluded from development and testing wherever possible.
- Hosting location and data residency follow your requirements.
Responsible AI
- Model choice. We select model providers and configurations that meet your data requirements, including options that don’t retain or train on your data.
- Grounding. Answers are grounded in approved sources and show references where it matters.
- Guardrails. AI agents get only the permissions they need; high-impact actions require approval.
- Evaluation. We test against agreed success criteria before launch and monitor quality afterwards.
- Audit trails. Automated steps and human decisions are logged so they can be reviewed.
Secure engineering
- Code changes go through version control and review.
- Secrets and credentials are kept out of code and stored in managed secret stores.
- Separate environments for development, testing, and production.
- Dependencies are kept up to date, with security updates prioritized.
- Access to client systems uses individual accounts and multi-factor authentication where available.
This website
frogcoder.com is served over HTTPS, sets no cookies, runs no analytics, and self-hosts its fonts. Hosting is provided by Vercel; contact form submissions are delivered by email. Details are in our Privacy Policy.
Compliance
We align our practices with recognised frameworks and your own policies. If your organization requires specific security questionnaires or assessments, we’re happy to complete them during scoping. Ask us through hello@frogcoder.com.
Report a security issue
If you believe you’ve found a security vulnerability in this website or anything we’ve built, please contact us through hello@frogcoder.com with details. Please give us reasonable time to investigate and fix it before disclosing it publicly. We appreciate responsible disclosure.
