Why I train

Javan Rasokat

My journey — from teenage bug hunting and game hacking through full-stack engineering into product security architecture, and finally being able to present my research at hacking conferences around the world (talks) — taught me I did not want to stay the application security team firefighter. I wanted paved roads: secure-by-design defaults that are easy to follow and hard to bypass, whether a human or an agent is shipping the code. That is the training I wish I had had as a practitioner years earlier — not an intro course, but the next level for people who already know how to find and fix bugs and want to design problems out, scale without staying the vulnerability firefighter, and keep that bar under today’s AI coding velocity.

I have always taught alongside the day job: Secure Coding as higher university lectures (e.g. DHBW University Karlsruhe), ISC2 certification trainings (e.g. CSSLP), and conference workshops built from current case studies from the work. In 2025 that became the DEF CON workshop on eliminating browser bug classes. In 2026 I delivered the Proactive Security Engineering core training at Black Hat USA — and every attendee gave it five out of five. The feedback confirmed what the classroom had already shown: the material lands when it stays practical, a little entertaining, and backed by both industry work and academic research (currently doing my PhD at the University of Bayreuth).

The topics keep moving — AI-assisted coding, platform-scale secure-by-design, latest browser features that help with defending — so the curriculum has to stay current. That is what pulled me from only being the practitioner into also being the trainer: try innovative approaches first at home, then bring them straight into the classroom.

— Javan Rasokat

Learning paths

Related workshops and trainings grouped by theme. Click a path again or All paths to clear.

Browse by interest

Browse by format

Courses

Proactive Security Engineering: Building Secure-by-Design Architectures That Scale

AI engineering now ships code faster than security reviews can keep up, and fixing bugs one at a time no longer scales. In this two-day hands-on training you exploit realistic flaws, trace them to their architectural roots, and rebuild with reusable secure-by-design controls that hold up at AI velocity, whether a human or an agent wrote the code.

Training · 2 daysIntermediateEnglish · GermanHeld at Black Hat USA 2026 · August 2026 · 2-day training
Course outline

AI coding assistants now ship features faster than security teams can review them, and they reproduce insecure patterns at the same speed. This two-day, hands-on training teaches how to evolve from vulnerability remediation to proactive secure architecture. Participants exploit realistic application flaws, including ones typical of AI-generated code, understand their architectural roots, and rebuild those systems using secure-by-design patterns (reusable platform defaults that make unsafe paths awkward) that scale across modern DevSecOps environments. Each module follows a repeatable cycle: exploit, test, defend, design pattern, re-test.

Coverage spans access control, input validation, secure configuration, design-from-the-start, browser security features, component governance, identity and secrets, SSRF prevention, and security logging. Where AI changes the picture, we cover guardrails for AI-assisted development: paved-road templates, policy-as-code checks in CI, and making sure AI-generated changes follow the same path. Industry case studies show which approaches held up. Labs use an enterprise-style training application.

Audience: Intermediate to advanced AppSec, platform security, architects, DevSecOps, and Security Champions

Own Your Vulnerability Harness: Prove the Exploit, Then Prove the Fix

Scanners and AI agents now produce findings faster than any team can triage. This two-day training teaches you to build and own a vulnerability harness: ingest, triage, sandbox, and fail-closed proof, then deepen it with reachability, runtime analysis, diff analysis, enrichment, and risk metrics that reprioritize on evidence.

Training · 2 daysAdvancedEnglish · German
Course outline

Finding vulnerabilities is no longer the scarce skill. Verifying that a finding is exploitable, in your environment, with evidence a machine cannot fake, is. This hands-on training shows you how to build a vulnerability harness you own instead of renting a black box.

Day 1 you build a four-stage harness (ingest, triage, sandbox, observed proof) that fails closed: a finding is "verified" only when deterministic, observable evidence shows the exploit worked, not when a model says so. Day 2 you extend it: exploitability and reachability analysis, runtime analysis, security-audit skills for agents, exploit verification per bug class, diff analysis of patches and commits, enrichment from advisories and threat intel, and dynamic reprioritization with risk metrics. For each approach we compare pros, cons, cost, and failure modes so you can choose what fits your stack.

All labs run on authorized or toy targets. You leave with a working harness skeleton, a set of fail-closed proof validators, a scoring model you can tune, and an adoption plan for wiring it into CI and agent workflows.

Audience: Intermediate to advanced AppSec, product security, and platform engineers who need to verify findings from scanners and AI agents instead of triaging them by hand

Attack & Defend Client-Side Browser Defenses: CSP, Trusted Types, and Fetch Metadata at Scale

Browser defenses are only as good as their weakest policy. In this one-day hands-on training you first break CSP, Trusted Types, and cross-origin isolation on a lab app, then lock each bug class down with the newest browser standards and learn how to roll the fixes out across hundreds of services.

Training · 1 dayIntermediateEnglish · GermanHeld at DEF CON 33 · August 2025 · 4-hour workshop versionHeld at BSides Las Vegas 2025 · August 2025 · 4-hour workshop versionHeld at German OWASP Day 2025 · December 2025 · 4-hour training version
Course outline

This is not an OWASP Top 10 or classic secure-coding course. It covers the latest changes in browser security features and shows how to attack and defend them - and how builders best scale the implementation across an organization.

You start on a lab app protected by modern browser controls. In the attack labs you bypass a weak Content-Security-Policy, get around Trusted Types, and leak data across origins despite COOP/COEP and Sec-Fetch-* checks. In the defend labs you fix each hole: nonce- and strict-dynamic-based CSP3, Trusted Types policies and sanitization, Fetch Metadata resource isolation, COOP/COEP, plus a first look at the Reporting API and emerging integrity policies. You re-test every fix under attack.

The second half is about scale: progressive enforcement from report-only to blocking, CI and organization-wide gates, and adoption metrics across hundreds of services. You leave with working policies, bypass test cases you can reuse, and a rollout plan - moving from patching individual bugs to eliminating whole bug classes.

Audience: Intermediate AppSec, platform security, architects, DevSecOps, Security Champions, and pen testers who want defence-by-design depth

Security Review Agents That Scale

AI coding assistants flood teams with multi-file diffs while human review depth collapses. This two-day training builds a shippable Agent-as-Judge: AISVS/ASVS-mapped rubrics, calibration against a human baseline, and a review-loop harness — not another scanner dashboard.

Training · 2 daysIntermediateEnglish · German
Course outline

You leave with a Review-Agent package: Judge prompts, rubrics, agreement log, and a team rollout plan. Day 1 establishes review quality baselines, threat-model → rubric backlogs, and Review Copilot Instructions. Day 2 configures a separate review agent (no self-review), calibrates against human findings, and runs a PR → review → fix → re-review → gate loop with measurable agreement.

Audience: AppSec architects, security engineers, staff/principal engineers, OSS maintainers with high PR volume. Bring your own AI coding tool with enough credits for labs. Intermediate. Distinct from the vulnerability-harness training (proving exploitability of findings) and from AI code-generation security tracks.

Audience: AppSec architects, security engineers, staff/principal engineers, OSS maintainers with high PR volume

Securing AI Code Generation Workflows

Teams already ship with Claude Code, Cursor, Copilot, or Codex — but scans, hooks, and CI were not designed for agent velocity. This two-day training secures the delivery path end-to-end and leaves you with a Pipeline Rollout package: generation instructions, hooks, gate matrix, and blueprint.

Training · 2 daysIntermediateEnglish · German
Course outline

Map Prompt → IDE/agent → commit → PR → CI/CD → runtime; build a Generation Security Harness (instructions, rules, hooks, MCP policies); implement generation/security/policy loops; and ship a gate matrix grounded in OWASP AISVS and ASVS.

Audience: DevSecOps and platform engineers, security champions, engineering leads adopting AI coding. Bring your own AI coding tool with credits. Intermediate; CI/CD familiarity assumed.

Audience: DevSecOps and platform engineers, security champions, engineering leads adopting AI coding

Secure Coding in the Age of AI: Attack, Fix, and Prove It

Learn how to find common web application security flaws, fix them with safer defaults, and prove each fix holds - skills that matter even more when coding assistants draft much of the code. Two-day beginner-friendly Secure Coding with hands-on labs.

Training · 2 daysBeginnerEnglish · German
Course outline

This two-day Secure Coding training walks through the OWASP Top 10:2025 topics that matter for everyday web apps. You attack a live training target, see how each issue works, then practice the secure coding fix - including deny-by-default and fail-closed paths - and close each topic with a negative test, a static check, and an acceptance bar so the control stays in place.

Most teams now draft code with assistants. This course still centers expert AppSec knowledge: for each topic you leave with a clear build rule and acceptance checks so you can ask for secure code, review AI-generated diffs, and refuse insecure defaults. Audience: software engineers and security champions new to web AppSec or refreshing fundamentals.

Audience: Developers and champions learning web AppSec fundamentals with attacker labs, safer defaults, prove-the-fix checks, and judgment for AI-assisted coding

Security Hackathon

I am lecturing in Software Engineering with a focus on Secure Coding. Using a mix of theory and practice, students learn about commonly exploited vulnerabilities in web applications and learn ways to exploit these, as well as measures to prevent or fix these vulnerabilities through secure software development.

TrainingBeginnerGermanHeld at DHBW Karlsruhe
Course outline

I am lecturing in Software Engineering with a focus on Secure Coding. Using a mix of theory and practice, students learn about commonly exploited vulnerabilities in web applications and learn ways to exploit these, as well as measures to prevent or fix these vulnerabilities through secure software development.

Diff Before CVE: Exploit Insight from OSS Patches – Agentic Weaponization?

Security teams still wait for a CVE number while exploit-relevant insight often appears earlier in open-source code diffs and quiet patch commits. This participate-friendly village walks real public fix commits, maps changes to exploit classes, and practices a reusable diff-to-hypothesis-to-class-map loop for research and triage.

Workshop · 2hIntermediateEnglish · German
Course outline

Security teams still wait for a CVE number while exploit-relevant insight often appears earlier in open-source code diffs and quiet patch commits. This village is a live, participate-friendly space about tools that watch those diffs and help surface exploit-class understanding before formal CVE publication.

During the event we run an approximately one-hour demonstration. We walk real public open-source fix commits, explain what changed, map that change to an exploit class, and discuss why early mitigations before the public patch is published mattered. Attendees then practice a reusable diff-to-hypothesis-to-class-map loop for research and triage. The village runs as a continuous laptop demonstration projected to a shared screen. Attendees may join at any point in the loop.

CSSLP Official Certification Prep (5-Day)

Five-day ISC2 CSSLP preparation through IT-Schulungen. The course covers all eight secure-software-lifecycle domains and prepares you for the CSSLP certification exam.

Training · 5 daysIntermediateEnglish · German
Course outline

This five-day course, ISC2 CSSLP — Certified Secure Software Lifecycle Professional, prepares you for the CSSLP certification exam. You build a working understanding of security across the software development lifecycle and how to integrate it from the start so applications are developed securely.

The course covers all eight CSSLP domains. It starts with secure software concepts and security design principles, then software security requirements: compliance, privacy, and data classification, misuse and abuse cases, and a Security Requirement Traceability Matrix (STRM). Architecture and design include threat modeling, the security architecture, secure interface design, and architectural risk assessment. Implementation covers writing secure code, analyzing code for security risks, and putting security controls in place. Testing covers security test cases, strategy, and plans. Lifecycle management includes secure configuration and version control and decommissioning software. Deployment, operations, and maintenance cover operational risk, secure release, and support during incident response. The secure software supply chain covers third-party software risk and checking pedigree and provenance.

Audience: Software architects, engineers, and developers; application security specialists; software program and project managers; quality assurance and penetration testers; software procurement analysts; and IT security managers

Get in touch

Questions about a course or a public venue date? I’ll reply by email.

I’m interested in