Trust & Compliance

Privacy, security and statutory compliance — built in, not bolted on.

cmplihr.ai is designed for organisations that handle sensitive employee, payroll and statutory data. Our trust posture is engineered to meet Indian regulatory expectations and recognised international standards.

Privacy by design

Your data, your control. India-first, India-only.

cmplihr.ai is engineered for organisations that must keep employee, payroll and statutory records under strict access, residency and accountability controls.

India-only data residency

Primary, replica and backup data remain in Indian regions. There is no cross-border processing of customer records.

Tenant isolation by default

Every row, query, file and AI prompt carries tenant scope. Default-deny authorisation, enforced at the data layer.

Data-subject workflows

Built-in workflows to handle access, correction and erasure requests in line with DPDP Act expectations.

Security posture

Defence in depth, with controls a regulator will recognise.

Encryption everywhere

Envelope encryption with tenant-scoped data keys, managed in a hardware-backed KMS and rotated on schedule.

  • AES-256 at rest
  • TLS 1.2+ in transit
  • Tenant-scoped DEKs
  • Scheduled key rotation
Identity & access

Least-privilege by default. Fine-grained RBAC/ABAC per module. SSO-ready and ready for just-in-time access elevation.

  • SSO via SAML/OIDC
  • Role and attribute policies
  • JIT elevation with approval
  • Session controls
Auditability as a primary surface

An immutable, exportable audit log is the system of record. Every user and system action is signed, attributed and time-stamped.

  • Signed event log
  • Who, what, when, where
  • SIEM-ready export
  • Retention controls
Secure operations

Environment separation, peer-reviewed change controls, and an external VAPT cadence designed for regulated workloads.

  • Isolated dev/stage/prod
  • Peer code review on every change
  • Quarterly external VAPT
  • Incident-response runbooks
Regulatory alignment

Aligned to the Acts and frameworks that matter.

We track and align our platform controls to the Indian statutory environment and recognised international information-security frameworks.

DPDP Act, 2023
Indian data protection

Aligned to the Digital Personal Data Protection Act. Consent, purpose limitation, retention and data-subject rights workflows are first-class.

Information Technology Act, 2000
IT Act & SPDI Rules

Reasonable security practices aligned to the SPDI Rules under Section 43A and supporting CERT-In advisories.

Indian labour codes
Statutory scope

Platform questionnaires, registers and notice flows track the Code on Wages, Industrial Relations, Social Security and OSH codes alongside state rules.

ISO/IEC 27001
On roadmap

An ISMS programme is in active build. Independent certification engagement on roadmap. Trust Letter available under NDA.

SOC 2 Type II
On roadmap

Designed against the Trust Services Criteria (security, availability, confidentiality). Independent attestation engagement on roadmap.

CERT-In directions
Operational

Log retention, time-sync and incident-reporting controls aligned to CERT-In April 2022 directions.

Certifications listed as "on roadmap" or "in progress" are active engagement areas and not yet awarded. Contact us for our current Trust Letter and roadmap dates.

controls × frameworks · coverage matrix
Control
DPDP
2023
IT Act
2000
CERT-In
Apr 2022
ISO 27001
Roadmap
SOC 2
Roadmap
India data residency
Primary, replica, backup in-region
Tenant isolation
Row-level scope, default-deny
Encryption at rest & transit
AES-256, TLS 1.2+, KMS DEKs
Identity & access (RBAC/ABAC)
SSO, JIT elevation, session controls
Immutable audit log
Signed, exportable, SIEM-ready
Data-subject workflows
Access, correction, erasure
Incident response & log retention
Aligned to CERT-In directions
Vendor / sub-processor register
DPA, residency, scope tracked
OperationalIn progressNot in scope
Responsible AI

AI assists. People decide. Every output is reviewable.

cmplihr.ai uses AI to structure, classify and draft. It does not file, submit, or send anything autonomously. A qualified human reviewer must approve every consequential output. We log model version, prompt context and reviewer identity on every draft.

No customer data is used to train third-party models
Tenant data never crosses tenant boundaries — including in AI prompts
Every AI draft records model, prompt context and reviewer
Customers can opt out of AI features at the tenant level
Enterprise diligence

Need our DPA, security questionnaire response, or pen-test summary?

Our security team responds to vendor diligence requests directly. Reach out and we'll share our Trust Letter, DPA template, and the current state of independent assessments under NDA.