Each vertical gets its own delivery playbook — the specific regulator that watches it, the recent letters and findings we're tracking, and the AI exposure patterns we see in that industry. Same SKUs across verticals. The compliance overlay changes by industry.
Senior associates pasting client P&L data into ChatGPT for variance analysis. Junior staff drafting client emails through AI without traceable provenance. The TPB's competence and confidentiality obligations apply to AI use — no grace period.
Advisers using ChatGPT or Claude to draft SOA paragraphs. Practice management software adding AI-powered "insights" that influence advice. Licensees demanding AI governance evidence; ARs having no answer.
Quote engines and renewal automation introducing AI features brokers don't control. Claims-side AI tools used without audit trails. Underwriters demanding AI usage evidence at binder renewal — brokers having no answer.
AI clinical scribes and transcription tools introducing accuracy concerns. Decision-support AI that may influence diagnosis. Practitioners personally responsible for AI appropriateness, data governance, and informed consent.
Estimators pasting tender bills of quantity into ChatGPT for variance checks. AI used to summarise subcontractor agreements. Major-project clients writing AI usage representations into head-contract terms.
Production-line vision systems from vendors who can't articulate AI governance. Predictive maintenance models on data the manufacturer doesn't have clear rights over. EU customers asking pre-supply AI governance questions Australian manufacturers can't answer.
Handa & Mallick is the leading Australian cautionary case. PI insurers introducing AI exclusions on renewal. We support modern, AI-curious firms — implicit pre-qualification on intake.
See the pageAnonymised but real — the findings, the numbers, and the response patterns are exactly what we see across mid-market and SMB engagements. Click through for the full write-up of each.
Power Automate + Anthropic Claude. Every incoming email classified into one of seven intents; drafts replies for the time-sensitive ones. Running in production.
Read the full case study →35-staff Australian SMB. Brief was cost optimisation — we delivered that and surfaced security findings the client didn't know they had.
Read the full case study →607 phishing emails to 560 recipients in 48 minutes. Hidden inbox rule, US IP rotation, classic BEC playbook. Detected, contained, and remediated through the night.
Read the full case study →ANONYMISATION FOLLOWS A WRITTEN CHECKLIST. CLIENT NAMES NEVER USED. NUMBERS AND FINDINGS PRESERVED. ASK FOR INDUSTRY-SPECIFIC EXAMPLES — info@evisent.com.au
AI Governance Bundle, Sprint and Build engagements are uniform across verticals — the compliance overlay is tuned to the regulator that watches you. Start where it makes sense.