Sunny Mahajan

Portfolio · New Delhi, India

Sunny Mahajan

Security Operations Lead · SOC, SIEM & Threat Detection

I run managed security operations for financial-sector clients, lead accessibility and compliance audit programmes, and build the tooling that makes a small SOC team faster: detection engineering, report generators and a human-in-the-loop AI triage engine.

Current: Practice Lead, Managed Security Operations at Codec Networks · Microsoft SC-200 · CRTA · AWS Certified

10+
years leading operations before and inside security
SOC
multi-client managed security operations on an enterprise SIEM
WCAG
2.2 accessibility audit programmes for financial-sector firms
5
security and cloud certifications, including SC-200 and CRTA

About

Operations discipline applied to security operations

I spent ten years running a business end to end: sales, service, suppliers, people and the reporting that kept it all honest. In 2025 I moved into information security at Codec Networks, and that background shapes how I work. A SOC is an operation. It needs documented procedures, named ownership, consistent investigation and reporting that a client can act on, not a feed of raw alerts.

Today I lead the Managed Security Operations practice and the Centre for Professional Excellence (CoPE), the firm's training arm. Day to day that means SIEM monitoring and tuning across several client environments, incident investigation and escalation, accessibility and configuration audits with evidence that stands up to scrutiny, and building the automation that lets a small team cover more ground. I hold the Microsoft SC-200 and CRTA credentials, so I test detections from the attacker's side as well as the console.

Security operations & detection

Running SIEM-based monitoring for multiple clients: log onboarding, rule tuning, use-case design, alert triage, threat hunting and escalation within agreed responsibilities.

Accessibility & compliance audits

Leading WCAG 2.2 A/AA digital accessibility testing programmes for financial-sector web and mobile applications, from Stage 1 assessment through revalidation to final reports.

Security assessment support

Web security audits, configuration reviews and vulnerability validation, with findings written for the people who have to fix them.

Automation & AI tooling

Designing and building the engines behind the work: alert triage with a local LLM, report generators, evidence capture pipelines and training platforms, all with a human approving every consequential step.

Experience

Where I have worked

July 2025 – present

Practice Lead, Managed Security Operations & CoPE

Codec Networks Private Limited · Gurugram, India · joined as Senior Information Security Engineer
  • Run SOC operations across multiple client environments in the financial sector: event monitoring, alert analysis, incident investigation and client escalation.
  • Engineer and tune enterprise SIEM deployments: agent rollout, log onboarding, detection rules, dashboards and use-case refinement.
  • Lead the digital accessibility testing programme for financial-sector clients, including Stage 2 revalidations, final report reissues and the report generation tooling.
  • Support security audits and compliance evidence; participate in Blue Team exercises; write technical reports and client presentations with actionable remediation.
  • Established and documented SOC policies, procedures and security guidelines; mentor analysts through CoPE learning programmes.
February 2014 – February 2024

Head of Operations

Shiv Shakti Automobiles · New Delhi, India
  • Managed day-to-day business operations across sales, service and administration, with responsibility for workflow, service quality and timely delivery.
  • Led and supervised operational teams; coordinated vendors, suppliers and internal departments.
  • Implemented process controls and operational improvements; maintained operational reports, compliance documentation and performance monitoring.

Projects

Selected work

Client work is described in general terms: no client names, data or findings are disclosed.

Managed SOC · 2025 – present

Multi-client managed security operations

Context
Multiple client environments in the Indian financial sector, monitored from a central SOC on an enterprise SIEM platform.
What I did
Onboarded agents and log sources, built and maintained the asset register, tuned detection rules, ran daily monitoring and investigation, and handled escalation with client IT teams. Alert-volume analysis showed that a small number of rules produced most of the high-severity noise, which focused the tuning effort.
Outcome
Documented SOC procedures, a known-good rule set and a volume baseline that guide tuning and the AI triage pilot.
Enterprise SIEMDetection engineeringMITRE ATT&CKIncident escalation
AI in the SOC · 2026

Human-in-the-loop alert triage engine

Context
Level 7+ alert volume was far more than an L1 analyst can read. The SOC needed help classifying alerts without handing decisions to a cloud model.
What I did
Designed and built a local triage engine: n8n workflows on PostgreSQL with an 8-billion-parameter model running on a local GPU. It classifies each alert as true positive, false positive or inconclusive, drafts the resolution note or client escalation email, and queues it for analyst approval. Known-good rules short-circuit routine events; guardrails, duplicate detection, decoding of obfuscated PowerShell and an append-only audit log are built in.
Outcome
51 unit tests and end-to-end scenario suite passing; 15 of 15 stability runs routed as expected. Nothing is closed or sent automatically: every recommendation waits for an analyst's decision. Currently a pilot.
n8nPostgreSQL 17Ollama · local LLMDockerJavaScript
Accessibility & compliance · 2025 – present

Digital accessibility testing programme

Context
Indian broking, wealth and asset-management firms must evidence WCAG 2.2 A/AA conformance for their websites and trading apps.
What I did
Lead testing against the 55 A/AA criteria on web, Android and iOS, including screen-reader passes; run Stage 2 revalidations and reissue final reports with updated statuses, health charts and recommendations; built the report generator and the browser-based evidence capture pipeline that produce consistent, auditable deliverables.
Outcome
Final reports issued across the programme, with every revalidation round tracked in a single stage-ordered archive and timeline.
WCAG 2.2Chrome DevToolsPlaywrightVoiceOver / TalkBackPython · python-docx
Web security review · 2025 – present

Security checks alongside accessibility audits

Context
Accessibility audits of public websites give a close look at every page, script and form.
What I did
Added a security pass to the audit routine: reviewing page scripts and third-party code, and checking that forms validate input correctly.
Outcome
Security observations are reported to the client promptly with evidence and prioritised by severity in the audit report.
Web security reviewJavaScript analysisEvidence capture
Training platform · personal project · 2025 – 2026

CYBERTRACE, a cyber investigation lab

CYBERTRACE home screen with a simulated enterprise network graph, an open incident and live events
The lab's home screen: a simulated enterprise graph, an in-progress credential-compromise incident and live events. All data is synthetic.
Context
New analysts learn triage fastest by investigating realistic incidents, but commercial ranges are costly. I wanted something a trainee can open in a browser.
What I did
Built a browser-based SOC training platform with its own SIEM-style query language, an evidence workspace covering 10+ log source types, an evidence graph, timeline reconstruction and MITRE ATT&CK mapping. Five scenarios cover account compromise, ransomware, data exfiltration, cloud account takeover and business email compromise, with scoring and an auto-generated incident report.
Outcome
Built as a single-file browser application; all scenarios, people and data are synthetic.
JavaScriptSingle-page appMITRE ATT&CKScenario design
AI agents · 2026

AI Employee OS: a governed digital workforce

Context
An experiment in delegating routine knowledge work to AI agents without giving up control.
What I did
Built six role-based agents (chief of staff, email, meetings, tasks, research, follow-ups) that act only through a registry of 15 tools. Anything external or irreversible requires an approval bound to a hash of the exact content. Includes a kill switch, prompt-injection handling, a secret scanner and a full audit trail.
Outcome
85 automated tests across unit, integration, API and security suites, plus an end-to-end journey; reasoning runs on the Claude API, a local model or a deterministic mock.
Next.jsClaude APIn8nPostgreSQLSecurity controls

Skills

What I work with

Security operations

SOC monitoringAlert triageThreat huntingIncident responseMITRE ATT&CKLog analysis

SIEM & detection

Enterprise SIEM platformsMicrosoft security operations (SC-200)Detection rulesUse-case designDashboards

Offensive & assessment

Red-team fundamentals (CRTA)NessusNmapKali LinuxWeb security review

Compliance & audit

WCAG 2.2Accessibility testingConfiguration auditCompliance evidenceSOC procedures

Cloud & AI

AWS fundamentalsAI practitionerLocal LLMs (Ollama)Claude APIn8n automation

Engineering

PythonJavaScript / TypeScriptNext.jsPostgreSQLDockerPlaywright

Credentials

Certifications and education

Microsoft

SC-200

Security Operations Analyst Associate

CyberWarFare Labs

CRTA

Certified Red Team Analyst · September 2026

Amazon Web Services

AWS Certified Cloud Practitioner

September 2026

Amazon Web Services

AWS Certified AI Practitioner

September 2026

NASSCOM

Certified Security Analyst

Version 3.0 · July 2025

Professional development

Forage · OSINTCon

Deloitte Australia Cyber Job Simulation, January 2026 · OSINTCon 2025

Education

MBA, Purchase & Operations Management

Swami Vivekanand Subharti University, Meerut · 2020

Let's talk

Open to security operations leadership roles, SOC and detection engineering work, and advisory on accessibility and compliance programmes for financial services.