Three engagements, each scoped to where your team actually is
Start with a Review that reads your current workflows. Move to a supervised Pilot when the brief makes the case. Sustain with ongoing Stewardship once your team is running independently.
← Back to HomeHow we approach AI integration for analytics teams
Every Kilauan engagement begins with reading — not presenting. Before we recommend anything, we look at a sample of what your analysts actually produce: ad-hoc SQL queries, dashboard outputs, and the stakeholder briefs that sit downstream of those dashboards.
From that reading, we identify where AI drafting support fits — and equally, where it does not. Not every workflow benefits from AI assistance. Some query patterns are too context-dependent; some briefs require domain knowledge that only your analysts carry. We write about those distinctions plainly.
Where we do recommend AI assistance, we design the integration around three principles: the analyst reviews everything before it leaves their desk, the data connection is read-only, and the engagement produces documentation your governance team can inspect.
Read before recommending
We review actual workflow samples before writing any recommendation.
Scope tightly
SQL drafting, dashboard summarisation, brief-drafting. Contained, specific use cases.
Deploy carefully
Read-only access, contained environments, IT-approved integration.
Document and sustain
Monthly sampling and audit briefs keep governance current as your team's use evolves.
Analytics Workflow Review
A short engagement where we sit with your head of data, one senior analyst, and a stakeholder representative. We read a sample of dashboards, ad-hoc queries, and stakeholder briefs, then write a brief on where AI drafting could carefully ease handling — typically SQL drafting support, interpretation summarisation, and stakeholder-brief drafting. The work respects existing data governance.
What the brief covers
SQL drafting patterns — where query complexity or volume makes a drafting assist worth considering
Dashboard interpretation — which dashboards generate high volumes of ad-hoc explanation requests
Brief-drafting bottlenecks — where analyst time is consumed by first-draft writing rather than analytical thinking
Technical setup
Warehouse access: Read-only credentials, scoped with your IT team
Environment: Contained, no data persisted between sessions
Supported platforms: BigQuery, Snowflake, Power BI, Looker
Drafting & Reading Pilot
A measured deployment of an AI drafting assist for SQL drafts, dashboard interpretations, and stakeholder briefs. The system drafts; analysts always edit before sharing. Integration with your warehouse is read-only and via a contained environment. Includes a four-week supervised period, training for two analysts, and a usage policy aligned with PDPA and your data classification.
Quarterly Stewardship
A monthly advisory engagement for analytics leaders whose teams already use AI drafting. Each month we sample outputs, audit data handling, and write a brief for your chief data officer. Includes alignment with PDPA, BNM-RMiT where relevant, and your data classification policy, plus a yearly written summary for your data governance committee.
What the monthly brief covers
Output quality sampling — a read of AI-assisted analyst outputs from the previous month
Data handling audit — review of access logs and usage against your data classification policy
Regulatory alignment notes — PDPA and BNM-RMiT flags where relevant
Which engagement suits your situation
| Feature | Workflow Review RM 1,200 |
Drafting Pilot RM 1,290 |
Stewardship RM 4,710/mo |
|---|---|---|---|
| Written advisory brief | ✓ | ✓ | ✓ monthly |
| Workflow reading session | ✓ | — | — |
| AI drafting environment setup | — | ✓ | Assumes existing |
| Analyst training (2 seats) | — | ✓ | — |
| Supervised deployment period | — | 4 weeks | Ongoing |
| Monthly output sampling & audit | — | — | ✓ |
| CDO brief each month | — | — | ✓ |
| Annual governance committee summary | — | — | ✓ |
| PDPA alignment documented | ✓ | ✓ | ✓ |
| BNM-RMiT consideration (FS clients) | — | — | ✓ |
Best for
Workflow Review
Teams exploring the idea of AI drafting. A complete standalone engagement that builds the internal case without committing to deployment.
Best for
Drafting Pilot
Teams with internal sign-off to trial AI drafting. The Pilot provides a structured, governed four-week test with training and documentation.
Best for
Stewardship
Teams already using AI drafting who need ongoing governance oversight, CDO reporting, and regulatory alignment maintained month to month.
Standards that apply to every piece of work we do
PDPA Compliance
All data handling follows Malaysia's PDPA 2010 requirements
No Data Retention
Client data is not stored between sessions or engagements
Written Output
Every engagement closes with a written brief, not a presentation
Analyst Review
AI drafts are reviewed and edited by analysts before use
Not sure which engagement fits? The Workflow Review is always the right place to start.
It gives you a written brief, a clear picture of where AI drafting fits your team, and the basis to make an informed decision about next steps.
Request a Quote