Skip to content
Back to Blog
·10 min read·compliance

Building a GRC Function in a Growing Indian Company

A practical guide to building a real GRC function for growing Indian companies: one risk register, one control set mapped to ISO 27001, SOC 2, DPDP and CERT-In.

BR

Bachao.AI Research Team

Cybersecurity Research

Check DPDP Compliance

Compliance risk for Indian SMBs

Non-compliance with the DPDP Act 2023 carries penalties up to ₹250 crore. This post explains what's at stake and what action to take.

Governance, Risk and Compliance (GRC) is not a document folder — it is the operating system connecting what leadership decides, what the business is actually exposed to, and what auditors and regulators can verify. For a 50-500 person Indian company, the minimum viable GRC function has four parts: a written risk appetite, a living risk register, one control set mapped to every framework you must satisfy (ISO 27001, SOC 2, DPDP Act 2023, CERT-In directions, and RBI/SEBI rules where applicable), and continuous evidence collection instead of a pre-audit scramble.

Most Indian mid-market firms treat GRC as a chore that appears once a year before an ISO surveillance audit or a customer security questionnaire. That is how GRC becomes spreadsheet theatre — a risk register nobody opens between audits, a policy PDF nobody reads, controls that exist on paper but not in production. This guide covers what GRC actually means, the minimum viable operating model, how to run a risk register people use, and how to map one control set to multiple frameworks without duplicating work four times over.

What GRC actually means beyond the acronym

Each word does different work.

Governance is who decides. It is the risk appetite statement, the policy set, the named owners, and the reporting cadence to leadership or the board. Without governance, teams fix whatever is loudest that week instead of what the business actually decided matters.

Risk is what could go wrong and how much it matters. A risk register is not a list of vulnerabilities; a vulnerability is a technical finding, a risk is the business consequence of that finding being exploited, weighted by likelihood and impact. Conflating the two is the single most common reason risk registers stop being used — they turn into un-triaged dumps nobody can prioritize.

Compliance is proving, to a named external party, that governance and risk management actually happened. ISO 27001 certification, a SOC 2 report, a DPDP audit trail — these are evidence artifacts a specific framework accepts as proof. Compliance without governance and risk underneath it is spreadsheet theatre: paperwork satisfying a checklist without reducing the chance of an incident.

ℹ️
INFO
A useful test: if you removed your compliance certificates tomorrow, would your security decisions change? If no, GRC is already working as an operating model. If yes, the certificates are the point and the controls behind them are for show.

The minimum viable GRC operating model for 50-500 people

You do not need a five-person GRC department at this size. You need one accountable owner (often a CTO, Head of Engineering, or a compliance lead reporting into the CTO/CEO) and a small set of recurring rituals.

  1. A one-page risk appetite statement, approved by whoever owns the P&L — what level of risk is acceptable in which areas (customer data, uptime, financial systems, vendor access), signed off, revisited annually.
  2. An asset and data inventory — what systems exist, what data they hold, who owns them. Nothing else in GRC works without this.
  3. A single risk register, reviewed monthly by the accountable owner, quarterly by leadership.
  4. One control set mapped to every framework the business needs — not four separate control lists per certification.
  5. Continuous evidence collection — logs, screenshots, ticket trails, and config exports captured as controls run, not reconstructed before an audit.
  6. A quarterly reporting cadence to leadership — top risks, control status, open findings, trend versus last quarter.
That is the whole loop — it runs continuously, not as an annual fire drill.
graph TD A[Set policy and risk appetite] --> B[Maintain asset and risk register] B --> C[Map controls to frameworks] C --> D[Collect evidence continuously] D --> E[Test and audit controls] E --> F[Report to leadership] F --> G[Feed findings back into risk register] G --> B style A fill:#1e3a5f,stroke:#3B82F6,color:#e2e8f0 style C fill:#1e3a5f,stroke:#3B82F6,color:#e2e8f0 style D fill:#1e3a5f,stroke:#3B82F6,color:#e2e8f0 style E fill:#1e3a5f,stroke:#3B82F6,color:#e2e8f0 style F fill:#1e3a5f,stroke:#3B82F6,color:#e2e8f0 style B fill:#1e3d2f,stroke:#10B981,color:#e2e8f0 style G fill:#1e3d2f,stroke:#10B981,color:#e2e8f0

The loop closes at the risk register, not the audit report. That distinction separates a working GRC function from a paperwork exercise — the audit is a checkpoint inside the loop, not the finish line.

Running a risk register people actually use

Risk registers fail for predictable reasons: too many rows, no clear owner per row, scoring nobody trusts, and no consequence when a risk sits open past its due date. Fix the mechanics before the content.

Keep the register to a manageable number of active risks — most functional registers at this company size run 20-40 live entries, not hundreds. Every finding does not deserve its own row; group related technical findings under one business risk statement.

Each row needs, at minimum: a plain-language risk statement (not a CVE ID), likelihood and impact scored on a simple 1-5 scale, a single named owner (a person, not a team), a treatment decision (accept, mitigate, transfer, avoid), a due date, and a status. A documented scoring rubric prevents every reviewer from grading on gut feel, which is the fastest way for a register to lose credibility.

💡
TIP
Put the risk register in front of the same leadership meeting every quarter, even when there is nothing dramatic to report. A register that only gets attention during a crisis trains the organization to treat risk management as a crisis response.
⚠️
WARNING
A risk register with items consistently past their due date is worse than no register at all — it is documentary proof, discoverable in an audit or breach investigation, that the organization knew and did not act.

Know your vulnerabilities before attackers do

Run a free VAPT scan — takes 5 minutes, no signup required.

Book Your Free Scan

Mapping one control set to many frameworks

The biggest efficiency gain in GRC at this stage is realizing that ISO 27001, SOC 2, DPDP obligations, CERT-In directions, and sector rules overlap heavily on the underlying control. Access control, logging, incident response, vendor risk, and encryption show up in nearly every framework — just phrased and evidenced differently. Build one control library and map each control to every framework clause it satisfies, instead of running parallel compliance projects.

ISO/IEC 27001:2022 organizes its controls into four themes — organizational, people, physical, and technological — with 93 controls total. That structure is a reasonable backbone for a unified library because most other frameworks' requirements map cleanly onto one of those four themes.

Control areaISO 27001:2022SOC 2DPDP Act 2023CERT-InRBI/SEBI (regulated only)
Access controlOrg/Tech controlsLogical accessFiduciary safeguardsAccess governance
LoggingTech controlsMonitoringBreach detection180-day log retention in IndiaLogging norms
Incident responseOrg controlsIncident managementBreach notificationReport within 6 hoursIncident reporting
Vendor riskOrg controlsVendor managementProcessor obligationsOutsourcing guidelines
Data protectionTech controlsConfidentialityCore Act + Significant Fiduciary dutiesLocalization/retention rules
Two entries need scoping care. CERT-In's April 2022 directions require reporting of specified cyber incidents within 6 hours of noticing them, plus 180-day ICT log retention in India — this applies broadly, not just to regulated financial entities. RBI and SEBI cyber security frameworks, by contrast, apply only to entities they regulate — banks, NBFCs, payment aggregators, stockbrokers, depositories, and similar intermediaries. Map RBI/SEBI-specific control language only where it genuinely applies to your sector.

Under the DPDP Act 2023, organizations classified as Significant Data Fiduciaries carry additional obligations, including a Data Protection Officer based in India, an independent data auditor, and periodic data protection impact assessments. Most growing companies aren't classified this way yet, but it's worth checking as your data footprint scales — see our DPDP compliance page for the fuller breakdown.

93Controls across 4 themes in ISO/IEC 27001:2022 (ISO)
6 hoursCERT-In incident reporting window from time of noticing (CERT-In, April 2022 directions)
180 daysMinimum ICT log retention required within India (CERT-In, April 2022 directions)
🛡️
SECURITY
If your company handles regulated data — payments, health records, financial transactions — confirm with legal counsel whether sector-specific RBI or SEBI directions apply before finalizing your control mapping. Applying them where they don't wastes effort; missing them where they do is a compliance gap.

Evidence collection: the difference between spreadsheet theatre and a real program

The single biggest tell that GRC has become theatre is evidence collected retroactively — screenshots taken the week before an audit, configs pulled from memory of "how it's usually set up," log samples cherry-picked to look clean. Auditors and enterprise security teams can tell the difference, and it affects both audit outcomes and deal cycles.

Continuous evidence collection means the artifacts an auditor needs — access review logs, patch records, vulnerability scan history, incident tickets, vendor due diligence records — accumulate automatically as controls run, timestamped and stored where they cannot be quietly edited later. Automated scanning earns its keep here: a recurring vulnerability assessment produces a dated, reproducible artifact every cycle, rather than a one-time PDF that ages out of relevance within weeks. Bachao.AI, built by Dhisattva AI Pvt Ltd, runs continuous automated VAPT for Indian SMBs and mid-market firms, feeding dated findings directly into a risk register instead of a static report left unopened until next year's audit — and where a CERT-In empanelled entity is specifically required, that is delivered with a CERT-In empanelled partner.

🎯Key Takeaway
GRC stops being spreadsheet theatre the moment the risk register, not the audit calendar, becomes the thing that drives what security work happens next. One control set mapped to every framework you need, evidence collected as controls run rather than reconstructed before an audit, and a risk register reviewed on a fixed cadence — that is the entire minimum viable operating model. Everything beyond that is scale, not a different discipline.

Where GRC effort actually goes in year one

For a company standing up its first real GRC function, year-one effort is rarely evenly split — most goes into risk assessment and control implementation before the program stabilizes into steadier evidence and audit-readiness work.

pie title Illustrative Year-One GRC Effort Split "Policy and governance" : 15 "Risk assessment" : 30 "Control implementation" : 35 "Evidence and audit readiness" : 20

Illustrative split — actual proportions vary by company size, control maturity, and how many frameworks are in scope.

A practical checklist to start this quarter

    1. [ ] Draft and get sign-off on a one-page risk appetite statement
    2. [ ] Build or update the asset and data inventory
    3. [ ] Stand up a single risk register with a documented scoring rubric and named owners
    4. [ ] Build one control library mapped to every framework you actually need
    5. [ ] Confirm whether RBI, SEBI, or Significant Data Fiduciary status applies to your business
    6. [ ] Set up continuous evidence capture instead of pre-audit collection
    7. [ ] Put risk register review on a fixed quarterly leadership agenda
ℹ️
NOTE
None of this requires a large team. A single accountable owner running these rituals consistently will outperform a larger team running an annual, disconnected compliance project.

GRC is not a certificate you earn once. It is the operating loop that turns leadership's risk appetite into daily control decisions, and those decisions into evidence that a framework — ISO 27001, SOC 2, DPDP, CERT-In, or a sector regulator — will accept. Build the loop once, map it to every framework you need, and audits stop being events and start being checkpoints on a program already running.

Frequently Asked Questions

What is the difference between a risk register and a vulnerability list?
A vulnerability list is technical findings from scans or audits. A risk register translates those into business consequences — what could happen, how likely, and how costly — scored, owned, and tracked to closure. Most GRC programs fail because they never make this translation.
Do we need separate compliance programs for ISO 27001, SOC 2, and DPDP?
No. These frameworks overlap heavily on underlying controls like access, logging, incident response, and vendor risk. Build one control library, map each control to every framework clause it satisfies, and run one evidence process instead of duplicating effort.
Does CERT-In's 6-hour reporting requirement apply to us if we are not a bank or financial company?
Yes, in most cases. CERT-In's April 2022 directions apply broadly to service providers, intermediaries, data centres, and body corporates operating in India, not just regulated financial entities. Confirm applicability with counsel based on your specific operations.
How often should a risk register actually be reviewed?
A functional cadence is monthly review by the accountable owner and quarterly presentation to leadership. Reviewing only before an annual audit is the most common reason risk registers become spreadsheet theatre.
What is a Significant Data Fiduciary under the DPDP Act 2023?
A classification applied to certain data fiduciaries based on data volume and sensitivity, triggering extra obligations including a Data Protection Officer based in India, an independent data auditor, and periodic impact assessments. Most growing companies aren't classified this way yet, but it's worth checking as your data footprint scales.
BR

Bachao.AI Research Team

Cybersecurity Research

AI-powered security research and threat intelligence from the Bachao.AI team. Covering the latest vulnerabilities, CVEs, and cybersecurity developments affecting Indian businesses.

Get cybersecurity insights for Indian SMBs

Weekly vulnerability alerts, DPDP compliance tips, and security guides. No spam — unsubscribe anytime.

We respect your privacy. Your email is never shared.

See if your business is DPDP Act compliant

Free automated scan — risk score in under 2 hours. No credit card required.

Check DPDP Compliance
Find your vulnerabilitiesStart free scan →