Security
Your data stays in your environment, under your control.
We build for firms that handle regulated, sensitive information. That shapes every decision we make. Your data lives in your own cloud, only two named people ever touch it, and you hold every control.
Who touches your data
Two named people. No one else.
Only the two directors, Chaitanya and Sanika Pandit, ever access your data. No employees, no freelancers, no subcontractors, no offshore teams. Your data is never handled by anyone you have not been introduced to.
A small, named, accountable team is easier to govern and audit than a large one. That is a deliberate choice, not a limitation.
AI assists. A person is accountable.
AI tools help us build and analyse. But they are never the accountable party. A named director reviews every AI-assisted output and is responsible for it. An agent cannot be held accountable for its actions. A person always can.
How your data is handled
Not all data is the same. We do not treat it the same.
Before any work begins, we classify your data and bind each type to a handling rule.
Secrets and credentials
Never leave a secured vault. Never sent to any AI tool under any circumstances.
Customer personal and financial data
Processed only inside your own cloud environment, on models running in your tenant. Never sent to an outside AI service.
Operational records
Processed only through tools that contractually guarantee zero retention and no training on your data, or inside your own environment.
Application code
Version-controlled with full history. Your data is excluded from code repositories entirely.
Inference is not training.
When a model processes data to produce an answer, it does not learn from or keep that data. Its weights are fixed. This is how inference works, not a policy promise. Your data is used to do the work, never to improve a model.
You stay in control
Every system we build. Every control you hold.
Deployed in your environment
Your data lives in your own cloud, in your regions. You control it. We do not hold it.
You hold every kill switch
We hold no standing production credentials. You review and approve changes before they go live. You can revoke our access instantly.
Build and live are separated
Development happens on synthetic test data. Real data lives only in production, controlled by you.
Dual-layer access control
Permissions are enforced in two independent places: the application and the database itself. A bug in one cannot widen access.
Audit trails and recovery
Every action is logged with who did what and when. Production runs automated backups with point-in-time recovery.
Hardened by default
Encryption in transit and at rest, multi-factor authentication on every account, least-privilege access, secrets kept out of code.
When an engagement ends, all client data is removed from our machines and data stores, our access is revoked, and you receive written confirmation.
Website security
We practice what we preach.
This website and the production applications we build are tested against internet.nl, an independent benchmark for web security posture.
Cybersecurity score
Based on internet.nl security test. Measures HTTPS, DNSSEC, security headers, and email authentication.
HTTPS & TLS
All traffic encrypted with modern TLS. HSTS enforced.
Security headers
X-Frame-Options, X-Content-Type-Options, Referrer-Policy configured.
DNSSEC
Domain name resolution cryptographically signed and verified.
Email authentication
SPF, DKIM, and DMARC configured to prevent spoofing.
Security and technology decisions are reviewed by a dedicated Technology and Cybersecurity Advisor through a risk and resilience lens. Our full Data Handling and Information Security Policy is available on request.