Agentic AI Security Consulting
Agentic AI Security Consulting — AI Agents for SecOps
Digvijay Parmar builds agentic AI security systems using LangChain and LangGraph — AI agents that triage, classify, investigate, and remediate across firewalls, NAC, cloud, logs, routing, and policy, with evidence-backed answers and audit-ready reports.
What agentic AI security means in practice
Agentic AI security is the next step beyond a chatbot. AI agents triage, classify, investigate, and remediate across your real security surfaces — firewalls, switches, NAC, cloud, logs, routing, and policy — using LangChain and LangGraph to chain reasoning, tool calls, and retrieval into a workflow that produces evidence-backed answers and audit-ready reports.
I build agentic security workflows that turn fragmented security data into answers an engineer can act on. The agents do not replace the engineer; they compress the 45-minute investigation into minutes and ground every recommendation in policy, logs, and telemetry.
Agentic systems I have built
FirewallIQ is a decision system for firewall policy operations. It ingests rules, topology, applications, owners, traffic evidence, and compliance context, reasons about them, and produces provably safe optimization recommendations under a strict change-governance workflow with Assist / Approve / Automate modes, multi-step approvals, simulation gates, and SHA-256 evidence packs. It cut policy review time by 60%.
The AI-Driven Security Investigation Platform uses LLMs and RAG to correlate firewall policy, Cisco ISE / NAC authorization logs, and BGP/OSPF routing data, automating investigation workflows and validating Zero Trust segmentation — cutting time-to-answer from ~45 minutes to under 5.
The Nexa Copilot concept is a security copilot that lets engineers ask natural-language questions across firewalls, switches, NAC, cloud, logs, routing, and policy, then receive grounded, traceable answers with citations to policy and telemetry through a secure hub-and-spoke model for sensitive backend controllers.
Governance is what makes agentic AI safe for production
An AI agent that can remediate firewall policy is powerful; an AI agent that can remediate firewall policy without approvals, simulation gates, and an evidence trail is a liability. Every agentic system I build enforces governance so nothing executes without approvals, a passed simulation, and guardrail checks — and every recommendation is explained with computed evidence, not opaque scores.
This is the standard agentic AI security has to meet to run in a Fortune 100 or financial environment. It is also the part most vendors leave out.
Book a free Agentic AI Standup
The Agentic AI Standup is a free 40-minute working session built specifically for teams deploying AI agents in security. Bring one real problem — an agent rollout you are unsure how to govern, a workflow you want to automate, or a control you need to prove. Leave with a diagnosis and a written summary in 24 hours.
What collaborators say
"Digvijay is very talented in Network Security and he comes up with different ideas to solve the problems, tracing an unknown network, understanding the situation and solving them. He introduces us to new ways to solve the issues and also makes our team aware of it."
— Vibhor Katiyar, Technical Operations Manager, Amazon Web Services
"While working with Digvijay on the same network engineering team but different projects, he was very responsive and with detailed accurate information every time. No matter if it was requesting where to locate documentation, identify a specific config on a device or explain how an appliance is working the way it is, you could always depend on Digvijay to get things done in a timely detailed manner."
— Matthew Calhoun, Manager of US Security Operations, Northern Trust
Stack & tooling
Related work
Frequently asked questions
- What is agentic AI security?
- Agentic AI security uses AI agents — built with frameworks like LangChain and LangGraph — to triage, classify, investigate, and remediate across firewalls, NAC, cloud, logs, routing, and policy. Digvijay Parmar builds these workflows in production at Point72 Asset Management.
- How are AI agents different from a security chatbot?
- A chatbot answers questions. An AI agent takes actions — querying live systems, correlating data, proposing changes, and executing them under governance. Digvijay's FirewallIQ platform runs in Assist / Approve / Automate modes with multi-step approvals, simulation gates, and SHA-256 evidence packs.
- Is agentic AI safe for regulated environments?
- Yes, when it is governed. Digvijay introduces AI agents with traceability, compliance alignment, and secure data pipelines — nothing executes without approvals, a passed simulation, and guardrail checks. This is the standard he builds to for financial-sector environments.
- What frameworks does Digvijay use for AI agents?
- LangChain and LangGraph for agent orchestration, with RAG for retrieval over policy, logs, and routing data, and Python for the integration layer. The architecture is designed to switch from mock data to live vendor APIs the moment credentials are supplied.
- What is an Agentic AI Standup?
- A free 40-minute working session where you bring one real AI security or Zero Trust problem and leave with a diagnosis, two or three concrete recommendations, and a written summary in 24 hours. Three slots open each week. No pitch, no deck.
Bring one real problem. Leave with a direction.
The Agentic AI Standup is a free 40-minute working session. You bring one real agentic ai security consulting problem; you leave with a diagnosis, two or three concrete recommendations, and a written summary in your inbox within 24 hours. Three slots open each week.
Book your free session See how it works