Live Cohort · Cloud Security + GenAI

Secure the
Cloud.
Build with AI.

Practitioner-led programmes in Cloud Security Engineering, Applied GenAI, and Pure LLM Engineering — from CSPM rule design to building your own agents, taught by a 19-year cloud security architect.

Cloud
×
GenAI
☁️
🛡️
🤖
🔮
🔍
476+ Security Rules Built
19 yrs Cloud Security XP
4 Learning Tracks
AWS Native Focus
Programmes

Four Tracks.
One Integrated Vision.

Each track stands alone — or stack them to build complete cloud security intelligence + AI engineering capability.

🛡️ Foundation

Cloud Security Engineering

Hands-on CSPM rule authoring, CIEM posture analysis, CWPP runtime controls, and compliance mapping across CIS, PCI-DSS, MAS TRM, and QCB frameworks on AWS.

⏱ 6 weeks 🏷 Live + async ⚡ AWS-native
🤖 Applied AI

GenAI for Security Engineers

Build LLM-powered security agents, design agentic remediation flows with LangGraph, integrate RAG pipelines for policy Q&A, and deploy MCP servers for tool-augmented reasoning.

⏱ 5 weeks 🏷 Project-based ⚡ LangGraph · MCP
🧠 Pure GenAI

LLM Engineering & Agents

No security context required. A standalone deep-dive into LLM fundamentals, prompt engineering, RAG, agent design, skill generation, LangChain, and LangGraph — for any domain.

⏱ 7 weeks 🏷 Cohort + self-paced ⚡ LangChain · LangGraph
🔮 Advanced

Predictive Security · FORESIGHT

Drift detection with time-series ML, anomaly scoring via vector embeddings, predictive risk surfaces, and real-world architecture from the FORESIGHT add-on built on QuestDB and Qdrant.

⏱ 4 weeks 🏷 Cohort only ⚡ ML · Vectors
Pure GenAI Track — No Security Context Required

LLM Engineering,
Start to Agent.

A complete, standalone curriculum covering everything from how transformers work to building production-grade multi-agent systems. Whether you're a developer, data engineer, product manager, or architect — this track takes you from first principles to shipping real AI applications using LangChain and LangGraph.

Topics Covered

LLM Fundamentals Transformers Tokenisation Prompt Engineering RAG Pipelines Vector Stores Embeddings LangChain LangGraph Agents Tool Use MCP Servers Skill Generation Memory Systems Multi-Agent Ollama · Local LLMs Fine-tuning Basics Evals
MODULE 01–02

LLM Fundamentals & How Models Think

Understand what LLMs actually are — attention heads, context windows, temperature, top-p, and why models hallucinate. Build a mini-LLM from scratch in NumPy to demystify the math.

Transformer architecture & attention Tokenisation, BPE, vocab Sampling strategies & decoding Context window management
MODULE 03–04

Prompt Engineering & Skill Generation

Master the craft of prompting — zero/few-shot, chain-of-thought, structured outputs with XML, and building reusable prompt skill libraries that your entire team can use.

System / user / assistant roles Chain-of-thought & self-critique Skill / SKILL.md pattern Structured JSON & XML outputs
MODULE 05–06

RAG — Retrieval Augmented Generation

Build a full RAG pipeline from ingestion to generation. Learn chunking strategies, embedding models, vector stores (Qdrant, Chroma, FAISS), and reranking for precision retrieval.

Document loaders & chunking Embedding models & similarity Qdrant, Chroma, FAISS Hybrid search & reranking
MODULE 07–08

LangChain — Chains, Tools & Memory

Go from LangChain basics to building complex chains with tool use, conversation memory, custom retrievers, and streaming responses. Ship a real Q&A application by end of module.

LCEL (LangChain Expression Language) Tool calling & function binding Conversation & summary memory Streaming & callbacks
MODULE 09–10

LangGraph — Stateful Agent Flows

Design and build stateful, graph-based agent workflows with LangGraph. Implement loops, conditional branches, human-in-the-loop interrupts, and parallel node execution.

State machines & graph nodes Conditional edges & routing Human-in-the-loop checkpoints Subgraphs & parallelism
MODULE 11–12

Multi-Agent Systems & MCP

Architect multi-agent systems where specialised agents collaborate, delegate, and share state. Build Model Context Protocol servers to expose external tools to any LLM orchestrator.

Supervisor & worker agent patterns Agent memory & shared state MCP server authoring Evals & agent benchmarking
Curriculum Highlights

What You'll Actually Build

No slides-only theory. Every module ships a working artefact — a rule, an agent, a model, or a dashboard.

01

AWS Security Rule Engine

Design and implement CSPM rules covering IAM, S3, EC2, VPC, Lambda — with real-time evaluation logic in Go.

02

LLM-Powered Triage Agent

Build a Claude-backed agent that reads findings, reasons about blast radius, and drafts remediation tickets autonomously.

03

CIEM: Identity Attack Paths

Map privilege escalation paths, overpermissioned roles, and cross-account trust chains using graph analysis.

04

RAG Policy Q&A System

Ingest your AWS Config, SCPs, and compliance docs into a vector store. Answer natural language policy questions in seconds.

05

Drift Detection Pipeline

Stream CloudTrail events into QuestDB, compute deviation scores, and alert on anomalous config changes before they become incidents.

06

MCP Security Toolchain

Build Model Context Protocol servers that expose nmap, IAM analyser, and Terraform linters as tools to an LLM orchestrator.

07

Compliance Automation

Auto-generate evidence packs for CIS, SOC 2, MAS TRM, and QCB from live AWS environments — no manual screenshots.

08

Predictive Risk Scoring

Train a lightweight model on historical finding patterns to surface tomorrow's likely misconfigurations today.

Who This Is For

Built for Practitioners

This is not an intro course. It's for engineers who already work in the cloud and want to operate at a higher level.

🏗️

Cloud Architects

Embed security intelligence directly into your design patterns and architecture reviews.

🔐

Security Engineers

Go beyond detection — build AI-augmented systems that reason, prioritise, and remediate.

⚙️

DevOps / Platform

Add security guardrails and preflight intelligence to your CI/CD and IaC pipelines.

🧑‍💻

Developers & Engineers

No security background needed for the GenAI track — build agents and ship AI features from day one.

📊

Risk & Compliance

Understand how AI-driven CSPM tools produce evidence and how to evaluate their outputs.

T
Thiru
Principal Cloud Security Architect · Kveesys

19 years working at the intersection of cloud infrastructure and security engineering. Builder of Perimeter Cloud — a multi-product CSPM/CNAPP platform with 476+ AWS security rules, an AI-powered preflight engine, and a predictive ML add-on targeting regulated markets in Singapore and Qatar.

476+ Production AWS security rules across CSPM, CIEM, CWPP, CNAPP
19 yrs Cloud security, Go engineering, and AWS-native architecture
3+ SaaS products shipped: Perimeter, InfraPilot AI, FORESIGHT
MAS·QCB Compliance frameworks: MAS TRM, QCB, CIS, PCI-DSS, SOC 2
Early Access

Next Cohort Opens Soon

Drop your email to get notified when enrollment opens — and receive the free Cloud Security × GenAI reading list.

or reach out directly on WhatsApp +91 93454 10863

Chat on WhatsApp