Four things,
done well.
We focus on four practices. Most engagements mix two of them. All of them end with something your team can keep using after we are gone.
AI Governance
A working governance model: decision rights, risk tiers, review paths. Not a binder — a single-page map your operators actually use.
- AI use inventory and tiering
- Decision rights matrix
- Model and vendor review checklist
- Executive briefing on emerging regulation
Policy
Internal policy that gets read. Acceptable use, data handling, vendor management — short enough to remember, specific enough to act on.
- Acceptable use policy — one page, plain language
- Data handling and redaction guidelines
- Third-party AI tool approval process
- Incident response playbook
Automation
Targeted automation and LLM pipelines. We build the first version ourselves — enough to prove the thing works before your team invests.
- Workflow audit — where hours are actually leaking
- Prototype automation in your stack, handed off with docs
- LLM-assisted pipelines: extraction, drafting, triage, summarisation
- Agentic systems where the ROI and risk profile support them
Training
Workshops and enablement. People leave with prompts, patterns and a shared vocabulary — not a certificate of attendance.
- Executive briefings — 60 minutes, high signal
- Two-day practitioner workshops on your real workflows
- Prompt libraries for legal, ops, sales, support
- Train-the-trainer for internal champions
Not sure where
to start?
The Survey is designed to be low-risk on both sides — two weeks, one written landscape document, and a prioritised recommendation.