Code assistance, documentation, incident management, and stakeholder communication — for developers, sysadmins, security analysts, and IT managers who want to move faster without introducing security risks.
Built for IT professionals across the full stack — from developers writing code to IT managers writing exec updates. Claude is already one of the most capable AI tools for technical work; this course teaches you to use it safely, efficiently, and without accidentally leaking credentials or customer data.
Use Claude for code review, debugging, and explaining code — with proper security sanitisation
Write runbooks, SOPs, and technical documentation faster and more consistently
Draft incident post-mortems, change requests, and stakeholder communications
Translate technical concepts into plain English for non-technical audiences
Apply credential security, NDB scheme, and ACSC Essential Eight principles to AI use
Leave with a Sprint-Ready Pack of working artifacts built on your own sanitised work — debugged code, a script or runbook, documentation, a stakeholder artifact, and a prompt library — ready to drop straight into the current sprint
Credentials, customer data, and production system details never enter Claude — at any point. All code examples use sanitised placeholders.
All code examples use sanitised placeholders. Production credentials and customer data stay out.
You are a senior Python developer reviewing code for bugs, performance, and readability. Context: This function queries a database to retrieve a user record. [Sanitised — no real DB credentials or hostnames] Task: Review for: (1) bugs, (2) performance issues, (3) readability improvements, (4) security concerns. Format: Annotated feedback by issue type. Provide a corrected version at the end. Note any issues requiring testing before production deployment. [Paste sanitised code below]
You are a senior Linux sysadmin. Context: I need a bash script for automated backups on an Ubuntu server. Task: Write a script that: (1) backs up [SOURCE_DIR] to /backups with a timestamp filename, (2) deletes backups older than 30 days, (3) logs all actions to /var/log/backup.log with timestamps, (4) exits with code 1 and logs an error if the source directory doesn't exist. Format: Commented script. Note any assumptions about system configuration.
You are an IT manager writing a stakeholder update. Context: We experienced a P1 incident last Tuesday. [SYSTEM NAME] was unavailable for 2.5 hours due to a misconfigured load balancer after a routine deployment. Root cause identified and fixed. No data loss. Task: Draft an executive update email for non-technical stakeholders. Format: Under 200 words. Plain English — no technical jargon. Acknowledge impact, explain cause briefly, confirm resolution, state what we're doing to prevent recurrence.
You are a senior SRE writing a post-mortem. Context: [SYSTEM NAME] experienced a database connection pool exhaustion incident. Duration: 47 minutes. Affected: [CUSTOMER SEGMENT PLACEHOLDER]. Root cause: connection leak introduced in v2.4.1 deployment. Detection: automated alerting at T+12 minutes. Task: Draft a post-mortem document. Format: Standard sections — Summary, Timeline, Root Cause, Contributing Factors, Action Items with owners, What Went Well, What Didn't. Blameless tone throughout.
Security hygiene with AI tools requires specific disciplines. We treat credential protection as non-negotiable.
API keys, passwords, tokens, private keys, database credentials, and service accounts must never be pasted into Claude. If Claude sees one, it will flag it — but the safest approach is never to paste it.
Before pasting code: replace real hostnames with [HOST], real IPs with [IP_ADDRESS], account IDs with [ACCOUNT_ID], and customer data with [CUSTOMER_DATA]. Always sanitise first.
Claude can identify common security patterns and anti-patterns, but it is a first-pass reviewer only. Dedicated SAST tools (Snyk, SonarQube, Semgrep) are required for production security assurance.
Claude's knowledge of CVEs, library vulnerabilities, and framework security advisories has a cutoff. Always verify security-relevant information against current NVD, vendor advisories, and official docs.
Using AI tools in your organisation should be assessed against your organisation's ACSC Essential Eight posture. AI model terms of service and data handling must be reviewed before use with customer data.
If sensitive personal information is accidentally pasted into Claude, assess your notification obligations under Australia's Notifiable Data Breaches scheme. Prevention is better — never paste customer data.
Ideal for whole development teams, IT departments, or mixed technical and non-technical cohorts.
Open to all IT and technology professionals.
Delivered at your office or remotely. Exercises use your tech stack and document types.
Book a public course or train your whole technical team in one day.