Straight answers for RIA principals & compliance officers

Are we still the adviser and the record-holder?

Yes — always. CloudLink is a technology and services provider, never the adviser. Your firm remains the regulated entity and the holder of its books and records. The architecture is deliberately built to keep us a vendor outside your regulatory perimeter, and your authoritative records stay in your existing, already-compliant systems by default.

How does this stay compliant with the Advisers Act?

Every obligation is mapped to a specific control: Books & Records (Rule 204-2) → an immutable, exportable audit trail; Marketing Rule (206(4)-1) → human approval and screening of all client-facing content, with records retained; Reg S-P (incl. the 2024 amendments) → data classified, minimized, encrypted, and isolated per firm, with an incident-response process. We provide a full security-questionnaire answer key your CCO can use directly.

Can the AI give advice or place trades on its own?

No. Agents draft and propose; a human approves before anything is sent, recorded, or acted on. Advice and client communications always carry a named human approver — the author of record — and every action is staged, approved, and logged on a tamper-evident audit trail.

What happens in an SEC exam?

The answer is evidence assembly, not archaeology. You get comprehensive surveillance, an audit trail that reconstructs exactly what happened and who approved it, and an obligation overlay that maps books and records, Marketing Rule, and Reg S-P obligations to the evidence that satisfies them. Good answers to a vendor-oversight review make your firm look good — that's the design goal.

Where does our client data (NPI) go? Is it used to train AI models?

No training on your data. Inference runs through enterprise channels under contractual no-training and retention limits, in-region. Your data is classified, encrypted in transit and at rest, and isolated per firm. Sensitive identifiers are redacted before anything leaves your boundary. We maintain a disclosed sub-processor list and a written information-security program.

How is our firm isolated from your other clients?

Each firm gets its own isolated data store and its own encryption keys, enforced structurally — not by a query filter that could be forgotten. "Cross-tenant data bleed" is a first-class, automatically tested failure mode that gates every release. We're building toward SOC 2 Type II covering exactly these systems.

What is the difference between modeled and measured numbers?

Modeled means built from cited assumptions and your calculator inputs. Sim-measured means demonstrated live on a simulated firm. Pilot-measured means measured in your environment against your baseline. We use exactly those three labels: Modeled, Sim-measured, and Pilot-measured.

What about AI "hallucinations" in client communications?

Three guardrails. Outputs are grounded and cited in your firm's own data. Every client-facing item is reviewed and approved by a human, who is the author of record. And automated evals screen for unsupported claims, Marketing-Rule issues, and data leakage before content ever reaches a person.

Will this replace my staff?

No — it augments them. The goal is capacity, not layoffs: each role gets an agent that does the busywork, freeing advisors for clients and ops for exceptions. Given the industry's advisor shortage and succession gap, the win is doing more with the team you have — and making those roles less administrative and more valuable.

Do we have to replace our CRM, custodian, or planning software?

No rip-and-replace. The digital twin overlays your existing stack — Microsoft 365, Wealthbox/Redtail, Orion/Addepar, eMoney/MoneyGuidePro, custodial feeds, Stripe — which remain your systems of record. We connect to what you have.

What happens when better AI models come out — or when we change our CRM or portfolio system?

The Ownership Line is how we build for that: the models are rented — routed through one gateway, chosen per task, swapped when a better one ships. Everything above the line is owned, by you: the versioned map of your firm, the workflow definitions, every approval your people ever made, the evidence of what the system actually did, and the controls an examiner would test. A model upgrade is a routing change. A vendor change is an integration project. Neither touches what you own — and what you own is what compounds.

How long until we see value? What does a pilot look like?

We start with a pilot on "quick-win" workflows (meeting notes, follow-ups, review prep, paperwork) that reclaim the most time fastest — value in weeks, not years. You see measured ROI before committing to the next wave (onboarding, billing, RMDs), then continuous functions (surveillance, rebalancing).

How does pricing work?

Two models. Platform: you own the software and accounts; we build and service it (a one-time conversion build + an annual platform subscription). Managed: we own the economics and bill you one metered invoice with a margin. Most firms start with a pilot and choose the model that fits their compliance posture. Founding-partner terms for the first five firms — 45% off the build and 50% off the Year-1 platform, always quoted against the published list — are shown directly in the ROI calculator, which computes the value against any price. We'll size it to your firm in a proposal.

What happens to our data if we leave?

No lock-in. You get a full export — canonical data, documents, twin-generated records, and the audit trail — in open formats, followed by verified deletion. Because your authoritative records live in your own systems throughout, leaving never puts your books and records at risk.

Is the ROI real, or just a sales number?

We model it conservatively from your own inputs, using published industry benchmarks, and show every formula and assumption. Industry research puts the administrative burden at 10–15 hours per advisor per week; we conservatively model roughly a third to a half of that as reclaimable: ~230–265 hours per advisor per year — about 5 hours a week. The pilot measures actual reclaimed time. We deliberately under-promise — credible numbers that survive a CFO's and a CCO's scrutiny are better for everyone.