Where we are

Three of six weeks done. Week 1 ran discovery sessions across the Group. Week 2 was a deep document review (88 documents across compliance, audit, finance, HR, structuring and reporting). Week 3 closed out the missing function sessions and brought the external vendor conversations into the picture, most notably the Quantios call. The remaining three weeks turn this body of evidence into a quantitative gap analysis, a target architecture, and a roadmap.

Hours sit at 58 of 80, which is by design. Stage 1 was planned front-heavy: discovery and document review carry the most CrossVal time, while the remaining work is synthesis, drafting and review of material already gathered. The envelope is a hard cap at 80 hours.

What we have covered

Three weeks of input, organised below by week so the cadence is visible.

Week 1: discovery sessions

Nine sessions across MHQ, Rethink and RAA.

Week 2: document review and remaining function sessions

The Tech Strategy folder went through a structured pass: 88 documents reviewed in total, presented as 71 rows in the working tracker. Coverage spanned compliance, audit, corporate secretarial, finance (MHQ internal, Rethink internal, external), HR, marketing, private clients, structuring, systems and reporting. Roughly 18 project-hours invested. Alongside the document work, the Rethink internal compliance and Rethink internal accounting sessions were held, closing the W1 outstanding list except for Barry's recruitment follow-up.

Week 3: external vendor + follow-up sessions

Three substantive conversations brought the picture together.

• Quantios platform call. Worked through the tiered functionality (Essentials, Standard, Premium) and mapped it against the Group's active headcount. Of roughly 193 staff, about 110 do not need a Quantios Core licence: they are either in non-system roles (admin, marketing, HR via BlueSky) or live on other systems (RAA on Inflo, Rethink accounting on Xero). On a default scope, that leaves ~82 licensed seats: 53 Essentials (corporate secretarial, structuring, compliance associates, private-client RMs), 24 Standard (compliance managers running KYC360, MHQ finance, systems/admin), and 1 to 5 Premium (final accounts, bank reconciliation, treasury). The seat count swings on one decision: whether Rethink client accounting stays on Xero (~55 to 60 seats) or moves onto Quantios (~95 to 100). Quantios includes generative-AI capabilities at all tiers and direct integrations with KYC360 RiskScreen, Laserfiche, BankClarity and others.
• Barry, recruitment follow-up. Closed out the session that was missed in W1. Sourcing is ~70% of recruiter time, conducted manually on LinkedIn. The team operates from a 4-year-old Excel tracker. Data quality is >50%, sufficient for migration. An ATS is warranted, but the team's own sequencing is: automation first (note-taking, call summaries, scheduling), ATS second. Recruitment data currently sits in a silo with no integration to central systems.
• Marketing materials review. Reviewed the latest materials shared by the digital marketing agency. The scope, cadence and reporting structure have been added to the inputs that feed the W4 gap analysis. The marketing function will be looked at on the same footing as every other workstream, with the team's own goals in mind.

Target architecture taking shape

The single most important insight from Weeks 1 to 3 is that the core problem is not the system of record. Quantios will solve the system-of-record problem if and when the migration lands. The problem that sits upstream of that, and which Quantios alone does not solve, is the intake layer: how data gets into the system in the first place. Today that work happens by hand, in email, in Excel, with re-typing between systems. Until that changes, any unified system of record will sit alongside a parallel Excel layer that the teams maintain to do their actual jobs.

The target architecture below puts the unified client tier at the centre. Every function reads from it and writes to it. The intake layer feeds it from outside (forms, KYC360, OCR, structured ingest). Quantios sits underneath as the system of record.

Where the biggest efficiencies will land

Three focus areas account for the majority of cross-Group time and the strongest evidence base from W1 sessions and the W2 document review. Each is shown below as a today-to-target flow. The classification and sizing happen quantitatively in Week 4.

What we are seeing

Six findings shape the rest of the engagement. These have moved from working hypotheses in the W1 update to evidenced conclusions across W1 to W3.

• The input layer is the unsolved problem, not the system of record. Quantios solves where data lives. It does not solve how data arrives. If capture stays as manual as it is today, the teams will keep maintaining Excel alongside any new system, and the consolidation benefit will not materialise. The intake layer is therefore being designed in parallel with the Quantios decision, not after it.
• MFID exists and works, but is not yet end-to-end. The Master File ID concept already binds Viewpoint records across MHQ and Rethink. Compliance officers use it to auto-populate workflows for shared clients. Extending it end-to-end across every system and every touchpoint is an engineering exercise, not a process change.
• Document drafting is the most consistent AI candidate. Named independently by Marcia (board minutes, compliance calendars), Omar (audit working papers), Anna (compliance summaries) and Keerthi (proposal sections). Confirmed in W2 across 29 ADGM incorporation drafts, multiple Rethink proposal templates, and the M/HQ board-minute templates. Volume is in the hundreds of recurring documents per year.
• Compliance is the single biggest concentration of cross-Group manual time. Onboarding, periodic reviews, regulatory returns and the Compliance Monitoring Programme together account for the majority of compliance team hours. The W3 Quantios call confirmed direct KYC360 integration as a standard ecosystem partner, which removes the largest manual chain (Excel-mediated screening cycles) from the current state.
• Cross-entity reporting is the biggest finance time sink. Every leadership report begins with Deepti's pre-step of mapping data between Viewpoint and Xero. Month-end close targets W1-2 but currently lands W6-8. The blocker is intercompany reconciliation, which a unified client tier removes structurally.
• Recruitment sits outside the central data fabric. Barry's team operates from a Excel tracker with no integration to central systems. Their own sequencing is automation-first (note-taking, scheduling), ATS-second. We are aligned with that view.

Quick wins shortlist

These items have firmed up since the W1 update. Each is now supported by at least two evidence sources (a function head plus a W2 document, or a vendor capability plus a W1 session). Sequencing and sizing happen in W4.

Where each workstream stands

Stage 1 runs as six sequenced workstreams feeding two synthesis outputs, as set out in the proposal. Status as of end of W3.

Synthesis outputs

The two synthesis outputs sit on top of the six workstreams.

Stage 1 deliverables progress

Three written deliverables plus the engagement review meetings. Status as of end of W3.

Engagement review meetings (the fourth named deliverable) are running on cadence. End-of-Stage review session sits in W6.

Path to delivery

Six-week engagement, three weeks done, three weeks ahead. Delivery in the week of 29 June to 3 July.

Plan for Week 4

The headline work of W4 is turning the qualitative picture from Weeks 1 to 3 into a quantitative one. Each gap surfaced in the engagement so far gets rated against three axes: type of problem, severity, and current time and cost. The output anchors the remaining 20% of Deliverable 1 and feeds the roadmap sequencing that follows in W5.

Three further pieces of work run alongside.

• API integration mapping. Section 2 sets out where data should flow. The integration map sets out how: which systems read from and write to the client tier, what the integration contracts look like, what already exists and what needs to be built. This is the major chunk of Deliverable 2.
• AI and tools assessment narrowing. The four-path framing (commercial software, AI tools, CrossVal-built, internal build) gets applied to the tooling candidates that have surfaced: drafting assistant for documents, OCR and structured intake for the input layer, Co-pilot and Claude governance framing. Each candidate ends the week with a recommended path and the alternatives documented.
• Synthesis A and B drafting. The transformation roadmap candidate sequence sharpens. The operating model firms up around what stays inside the Group post-Stage 2 versus what is better sourced.