Spot it, stop it, work out what happened
Say an employee's account starts downloading the customer database at 3 a.m. Three kinds of tool deal with it, one after another.
Box 1 spots the download among millions of records. Box 2 runs the response automatically. Box 3 works out afterwards what the attacker did. Tap a title in the diagram to jump to its companies.
SIEM: one record of every alert 3 companies
A SIEM (security information and event management) collects the logs and alerts from every product into one searchable record. It also learns how each user and machine normally behaves, and flags what does not fit.
SecuronixSecuronix sells a SIEM that combines the record, the behaviour checks and the automated response. It has engineers in Bengaluru and Pune. Since 2025 it has also owned ThreatQuotient, which collects information about known attackers.Bengaluru · PuneHiring now
GuruculGurucul, with engineers in Pune, started with checks on how users behave, and later built a SIEM around them.Pune
VehereVehere, in Kolkata, records the traffic inside a network, which gives the SIEM the network half of the evidence.KolkataSee alsoNetwork security, firewalls and SASEAutomating the response 4 companies
A playbook is the list of steps the team takes for each kind of alert: look up where the download is going, cut the laptop off the network, disable the account, open a ticket and tell the right people. Done by hand, this takes an hour; run automatically, it takes seconds. This is called SOAR.
SwimlaneSwimlane sells Turbine, a general tool for automating the security team's work. It has engineers in Hyderabad.
CywareCyware, from New York with engineers in Bengaluru, combines three products: information about known attackers, automated playbooks, and a way to share that information between companies.Investigating after an attack 2 companies
After an attack, investigators collect data from phones, computers and cloud accounts, and work out what the attacker did. This is called digital forensics.
What's changing
AI now does the first look at each alert.
A security team receives far more alerts than its analysts can read. AI models now do the first look at each alert: they collect the related records, decide whether it matters, and write a summary, so analysts spend their time on the few that are real.
Companies: ingrid's company map. Hiring signals: postings feed of 5 Nov 2026.



