Manual, evidence-led testing of your applications, infrastructure and APIs. Findings you can act on, not a scanner export with a logo on it.
Most organisations buy a penetration test because someone asked for the report, not because they want to know the answer. That shapes everything: the scope gets narrowed until it is cheap, the test runs against a staging environment that does not match production, and the report arrives with forty findings sorted by a scanner's severity rating rather than by what an attacker would actually do.
The useful version is different in three ways. The scope includes the things you are afraid of rather than the things that are easy to test. Findings are rated by what they let an attacker reach in your environment, not by CVSS in the abstract. And the report distinguishes between what was tested and found safe, what was tested and found broken, and what was never tested at all, because the third category is where the next breach usually comes from.
A clean report that does not say what it did not cover is not reassurance. It is an unanswered question.
Anyone with an annual VAPT obligation under RBI, SEBI, IRDAI or CERT-In direction, and teams who want real assurance before a launch.
Not a list of everything possible — the things that come up again and again, in roughly the order they come up.
A vulnerability assessment enumerates weaknesses, largely with tooling, and is broad and shallow. A penetration test attempts to exploit them and chains them together, and is narrow and deep. VAPT as sold in India usually means both, but the proportions vary enormously between providers. Ask what share of the engagement is manual.
Two to six weeks for most scopes, including reporting and a retest. A single web application is at the short end; an estate with several applications, an API and an internal network is at the long end. Anyone quoting a fixed week for an unseen scope is quoting a scan.
Not with a competent tester and an agreed scope. Denial-of-service testing is excluded by default and only included if you ask. We agree testing windows, rate limits and an escalation contact before anything starts, and we stop immediately if something behaves unexpectedly.
Production where it is safe, because staging rarely matches it. Where production testing is genuinely too risky we test staging and say so in the report, including what that means for the confidence of the findings.
We provide a report and, where a customer or auditor needs one, a letter of attestation stating the scope, dates and methodology. Be cautious of anyone selling a security certificate for a penetration test; no accredited scheme issues one.