1/40
Looks like no tags are added yet.
What does a GRC analyst actually do?
Protect the company by balancing three areas:
Governance sets strategy and policies
Risk Management figures out what could go wrong and what to do about it
Compliance ensures legal and regulatory obligations are met
What is SOC 2?
AICPA audit framework proving a company protects customer data. Built on trust services criteria: security, availability, processing integrity, confidentiality, privacy.
Type 1: design at point in time
Type 2: Operating effectiveness over a period of time usually 6 to 12 months
Security is the only one required for SOC 2
Walk me through SOC 2 readiness
Scope First (type 1 or 2, criteria, system boundary)
Gap analysis across criteria categories against current state
Risk rank gaps and build a phased roadmap (5 to 8 weeks: Quick wins, Core processes, policies)
Draft Core policies (access control, incident response, vendor management)
Independent CPA audit and management assertion
Walk me through a NIST gap assessment
Scope and purpose (e.g. NIST 800–53 alignment)
Pick impact baseline (moderate) and tailor.
Reuse existing work (cross-reference SOC 2 Controls)
Rate families (satisfied, partial, gap), rank risks, build 12 week roadmap
Deliver system security plan (SSP) and POA&M
SOC 2 versus ISO 27001
SOC 2 Examines controls against trust service criteria within report scope. ISO 27001 specifies require requirements for an information security management system. (ISMS)
NIST CMF vs ISO 27001
CSF organizes/communicates security outcomes.
RMF is the process for categorizing, selecting, implementing, assessing, authorizing, and monitoring controls.
Framework vs. regulation
Framework: voluntary guidance (NIST CSF, ISO 27001)
Regulation: legally binding with penalties (GDPR, HIPAA)
Policy, standard, and procedure
Policy: high-level mandatory rule (What and why)
Standard: measurable require requirement enforcing policy
Procedure: step-by-step instructions for execution
3 Lines of Defense
Business operations, owning risk
Risk and Compliance setting policy and overseeing
Internal audit providing independent assurance
Segregation of duties
No one person controls a sensitive process from start to finish (eg approver vs Processor) To prevent fraud and error errors
KPI vs KRI
KPI: how well hitting objectives
KRI: early warning that risk is rising
BCP vs DRP, RTO, RPO
BCP: keeps business running through disruption
DRP: restores IT systems/data
RTO: maximum allowable downtime
RPO: maximum allowable data loss measured in time
Internal vs external audit
Internal: ongoing employee checks
External: independent, third-party opinion (CPA firm for SOC 2)
Vulnerability assessment vs pentest
Vulnerability assessment: automated scan listing known weaknesses without exploiting
Pentest: active exploitation to show real impact
Data classification
Public, internal, confidential, restricted. Determines storage, sharing, and disposal.
Incident response lifecycle (NIST 800–61)
Prepare
Detect and analyze
Contain, eradicate, recover
Lessons learned
Shared responsibility in the cloud
CSP: Security OF the cloud ( Infrastructure, hardware, data centers.)
Customer: Security IN the cloud (IAM, data classification, patches, config.)
What is least privilege?
Users get only the access needed to do their job, reducing compromise impact
Encryption basics
At rest (discs/databases) and in transit (TLS). Symmetric (fast, bulk data) versus asymmetric (key pairs). Key management is crucial.
Tell me about a mistake you made
In the first version of Risk Whisperer, I trusted the AI to produce control names and definitions, and some wrong or overlapping like CC 6.1. That was on me for not verifying the source first. I fixed it by building a verified look up table of AICPA criteria into the code and adding a second model to check mappings, flagging week fits for human review.
The lesson learned was that Accuracy comes from ground truth, sources, and guard rails, not confident sounding AI output
What if someone doesn’t respond to a request?
My short-term routine would be to send clear reminders stating what’s needed, why, and the deadline.
My long-term upgrade would be treat repeat manual chasing as a signal to automate evidence collection via API’s and scripts.
Are you comfortable with following up repeatedly?
Yes, but I would do it respectfully. I don’t let important items disappear, because I know clear follow up is part of good risk management.
how do you deal with a difficult person?
stay calm and ask what they actually need
find the goal we share
agree on next steps
and confirm in writing
For example, when a client wanted to sign off, prematurely on a wallet control, I showed them how an auditor would fail it, turning it into a successful 3 day sprint.
How do you handle a disagreement?
I share my view with evidence, listen to theirs, and defer to whoever owns the decision. Once it’s made, I back it fully.
How do you handle tight deadlines or pressure?
I break the work into the next smallest steps, protect the one thing that can’t slip, and tell people early if something will.
How do you ask for feedback?
I ask for specifics, apply them immediately, and follow up to close the loop.
For example, when asking if Risk Whisperer AICPA descriptions were clear enough for non-technical folks, which led to building The Plain Talk/Tech Talk toggle.
How do you learn something quickly?
I start with the primary source like frameworks or docs and then build something small with it to force, hands-on repetition and expose knowledge gaps.
Ex: like writing a python script to poll GitHubs API’s for continuous compliance
What do you do if you don’t know the answer?
I’m transparent about what I don’t know, I find the ground-truth answer in policy/frameworks or by talking to the process owner, and document it so the next person never has to ask
How do you handle telling someone no?
I don’t just say no. I explain the risk in business terms and show what an auditor would flag.
Then I offer a way to get what they need, like a compensating control or a short fix plan.
If it can’t be done safely, I put it in writing and the person who owns the risk decides.
Ex: The SOC 2 sprint, My client wanted to sign off on a control before the Wallet documentation was ready. I showed them how an auditor would fail it, and we fix the gap in three days.
How is your attention to detail?
I verify against ground truth, and I never trust a check until I’ve tested it.
For example, my evidence collector once showed a false pass on its get held branch protection check, because it passed on any protection at all.
I fixed it so it only passes when all 3 rules are in place. I use that same habit in Risk Whisperer where checking against the source. PDFs caught 5 drifting SOC 2 definitions.
The Takeaway is a check that passes the wrong thing is worse than no check at all.
How do you handle confidential information?
I share it only with people who need it, store it where access is controlled, and don’t discuss it outside of work.
If I’m unsure whether something is sensitive, I treat it as sensitive.
Tell me something that’s not on your résumé
i’m really hands-on. my curiosity pushes me to understand how things work, and my persistence drives me to fix them or make them run better.
How do you work with tech teams?
I speak their language by focusing on code, config, and data structures. And help build automated guard rails that make compliance friction list for them.
How would you improve a process?
I would find the step wasting the most time, measure it, remove or automated, and confirm it works
What makes evidence good?
evidence that is tied to the specific control, the correct period, the source system, and a reviewer, with a clear chain of custody.
That’s what my evidence collector does: timestamp, source, control, mapping, method documented
Why is system generated evidence better than screenshots?
It’s time stamped, pulled from the source, repeatable, and harder to alter than point in time screenshots
How do you do a vendor risk review?
Tier vendor by criticality
Review security docs/contracts
Identify gaps
Track remediation/risk acceptance
Ongoing monitoring
How are AI agents used in compliance operations?
Use agents for repetitive work (Collecting evidence, drafting control, tax, flagging, gaps)
Humans approved anything going to an auditor
AI need to ground-truth guardrails
how do you check AI output?
ground truth source
Verified look up tables overwriting AI drift
Second Model checking mappings
Week fits flagged for human review
What is third-party or vendor risk?
It’s the risk introduced by a vendor, supplier, or service provider, that handles company data, connects to systems, supports operations, or has access to sensitive information
You assess controls, access, data, handling, incident response, In contractual requirements