Microsoft Sentinel
Incidents with their accounts, hosts and IP entities, triaged as they arrive
A regulated institution has to watch its systems around the clock, report an incident within hours and show an auditor why each decision was made. Cyvion does the first-line work at any hour and keeps the record.
Early access. Built and run by Infovion Technologies, Pune.
Cyvion, a security operations platform from Infovion Technologies, helps banks, NBFCs, brokers and insurers in three ways. It triages every alert from the institution's SIEM and EDR at any hour and records the evidence for each verdict. It tracks exposure across networks and web applications between formal audits. And it drafts incident reports from the case record, with the timeline already in place, for reporting clocks such as CERT-In's six hours. Cyvion can be installed inside the institution's own data centre. It supports these obligations; it does not by itself make an institution compliant, and it holds no certifications yet.
Updated 17 November 2026.
Sign-ins, process starts, firewall blocks, cloud changes, reported emails. Thousands of events an hour, from every system you run. No team can read them all.
0 events
Cases that need a decision at the top, everything else already closed with its reasons. This is the console's overview on the demonstration tenant.
6 cases need a decision
The Cyvion console with its demonstration tenant. Names, addresses and figures are sample data.
Keep your SIEM and EDR. Cyvion takes their alerts, plus raw activity from identity, cloud, code and endpoints, and triages all of it the same way.
Incidents with their accounts, hosts and IP entities, triaged as they arrive
Alerts from Defender for Endpoint, Identity, Office 365 and Cloud Apps, with process evidence
Enterprise Security notable events, or the results of any search you choose
Endpoint and identity alerts with host, user, command line and file hash
Sign-ins, directory changes, mailbox rules, file activity, and two response actions
Login, admin, OAuth token and Drive events
CloudTrail management events and GuardDuty findings, per region
Organisation audit log, secret-scanning and code-scanning alerts
Account audit log and WAF firewall events per zone
Windows and Linux: logons, processes, sudo, SSH, accounts, software inventory
Post alerts as simple JSON from any SIEM, EDR or SOAR step
Cyvion's own detections, mapped to MITRE ATT&CK. Alerts from your SIEM and EDR are triaged on top of these, whatever they detect.
42 detection rules covering 33 techniques across 11 tactics. Choose a technique to see the rules behind it.
Credential access
A planning summary, not legal advice. Read the regulator's current text; several of these directions have been amended.
| What is expected | Where it comes from | How Cyvion helps |
|---|---|---|
| Report specified cyber incidents within six hours of noticing them | CERT-In Directions, 28 April 2022 | The case holds the timeline and evidence; a CERT-In report is drafted from it |
| Keep ICT system logs for a rolling 180 days | CERT-In Directions, 28 April 2022 | Your SIEM or log store remains the archive. Cyvion keeps cases, verdicts and evidence |
| Notify the regulator of incidents within its own deadline | SEBI CSCRF for regulated entities; RBI directions for banks and NBFCs | Time-stamped case record and report draft for each incident |
| Run a security operations centre or continuous monitoring | SEBI CSCRF; RBI Cyber Security Framework and IT Governance Master Direction | Every alert triaged, with people approving responses |
| Carry out periodic vulnerability assessment and penetration testing | RBI and SEBI frameworks | Continuous exposure scanning between cycles. It does not replace an empanelled auditor |
| Take reasonable security safeguards and report personal data breaches | Digital Personal Data Protection Act, 2023 | Detection, an incident record, and an audit log of access and actions |
Cyvion is in early access, with no certifications yet. Detection, triage, approvals, exposure scanning, agents and single sign-on are built and tested; the native connectors and the language-model analyst are being validated on live data.
Yes. Cyvion can be installed on the institution's own servers, with events, cases and reports stored there. An on-network instance is also what allows scanning of internal address ranges.
No product does that by itself. Cyvion supports specific obligations: continuous monitoring and triage, evidence for decisions, incident reports and exposure tracking. Policy, governance, audits and the regulator's other requirements remain the institution's work.
CERT-In's Directions of 28 April 2022 require service providers, intermediaries, data centres, body corporates and government organisations to report specified cyber incidents to CERT-In within six hours of noticing them, and to keep ICT system logs for a rolling 180 days.
With an on-premise installation and the playbook engine, no data leaves your network. If you enable the language-model analyst, the content of each case is sent to Anthropic's API for analysis.
Yes, through alert intake over a webhook. Alerts are sent in a simple JSON format, which the automation rules in Microsoft Sentinel, IBM QRadar, Splunk and similar tools can produce. Native connectors for Sentinel, Splunk and CrowdStrike are built and in validation; QRadar is not covered yet.
That is where first-line automation helps most: alerts are investigated at any hour and a small team reviews the cases that need a decision. The institution still needs named people to approve responses.
No. Empanelment applies to information security auditing organisations. Cyvion is a software product in early access and holds no certifications yet; an independent penetration test is planned before general availability.
Sources: CERT-In Directions under section 70B(6), 28 April 2022, SEBI: Cybersecurity and Cyber Resilience Framework (CSCRF), August 2024, RBI Master Direction on IT Governance, Risk, Controls and Assurance Practices, 2023, MeitY: Digital Personal Data Protection Act and Rules
Tell us which SIEM and EDR you run and roughly how many alerts a day. We reply within one business day and arrange a demonstration on your use cases.