No. SleekAI's system prompt is configured to share philosophy and process only - never specific security recommendations, retirement income projections, or individualised tax-aware sales strategies. Those are advice questions that depend on the prospect's full facts and circumstances, and they belong in front of a CFP at your firm with the relationship and the disclosure framework in place. The bot frames every reply as general information explicitly.

Yes. Add ADV Part 2A, ADV Part 2B, and Form CRS as PDFs or pages and SleekAI links them automatically when prospects ask about fees, conflicts, or how the firm operates. Many firms also configure proactive disclosure surfacing - the bot mentions Form CRS at the end of any fee-shaped exchange - which mirrors the disclosure flow a human intake coordinator would follow and is what compliance reviewers expect to see.

Yes. As long as your equity-comp service page lives on your site, SleekAI explains your process for ISOs, NSOs, RSUs, ESPP, and 10b5-1 plans at a high level. It will describe what your firm typically does for concentrated-position management - diversification planning, tax-aware sales, exchange funds, charitable strategies - without recommending any specific approach. The actual planning conversation happens with the CFP and the prospect's full picture in view.

Logs live in your WordPress database under your control. Most firms export them to compliance archives (Smarsh, Global Relay, MyComplianceOffice, or an internal system) on a recurring schedule via webhook. That keeps the chat transcripts inside the same archival workflow as your email and other written communications, which is exactly what a compliance audit will ask to see when they ask about your AI-assisted intake.

Yes. Multibot lets you run a HNW individual bot on consumer-facing pages (retirement, equity comp, business owner) and an institutional bot on family-office or endowment pages, each with appropriate framing, disclosures, and minimums. Institutional intake asks different questions and references different disclosures than HNW individual intake; the separation by display condition keeps each conversation tightly scoped.

Claude Sonnet tends to follow strict 'no recommendation' system prompts most reliably across long conversations, including under pressure when a prospect repeatedly asks for specific advice. GPT-4o is also strong and many firms run it. Smaller or older models drift more, which on an RIA site is exactly the failure mode you cannot tolerate. Bring your own key for whichever provider you prefer; the model is a settings change rather than a re-build.

Yes. Business-owner intake is its own practice area on most wealth firms, with questions around eventual sale, valuation, key-person planning, qualified retirement plans, and entity-level tax strategy. The bot can describe the planning framework your firm uses for owners and route to the specialist who handles owners specifically. As with everything else, it shares process and defers the actual recommendations to the discovery meeting.

The bot captures source-of-wealth conversationally - retirement, business sale, inheritance, equity comp, or other - as part of the qualifying flow. That information lands in the conversation log and then in your CRM via webhook so your AML and onboarding processes have the data they need at the start of the relationship rather than discovering it in a second or third meeting. The bot does not run AML screening itself - that is a back-office function - but it captures the upstream facts that the screening relies on.