No. The system prompt is configured to explain how claiming works at a general-education level, the tradeoffs between early, full-retirement-age, and delayed claiming, but never to recommend a specific age for a specific prospect. That recommendation depends on health, spouse, other income, and household goals, which a chatbot has no business assessing. Advice-shaped questions get routed to a credentialed CFP at your firm with the conversation summary attached.

Yes, at a general-education level. The bot walks visitors through Medicare Parts A through D, the initial enrollment window, late-enrollment penalties, and how IRMAA surcharges work for higher-income retirees. It uses your own guide and your service pages as the source. It will not, however, recommend a specific Part D plan or carrier, that is a benefit-selection decision that belongs with a credentialed agent or your CFP.

Yes, but only at the level of explaining the mechanics. The bot describes the relationship between conversion timing, tax brackets, IRMAA cliffs, and future RMDs, and it notes that Roth conversion modelling is included in your planning engagement. It does not advise on whether a particular prospect should convert, that is a personalised tax question that has to come from a CFP after reviewing the full picture, including state tax.

Yes. Upload ADV Part 2A, Part 2B, and Form CRS as PDFs or pages, and SleekAI links them automatically when prospects ask about fees, conflicts, or how your firm operates. Many retirement-planning firms configure the bot to surface the disclosures proactively at the end of any fee discussion, mirroring the disclosure flow a careful human intake coordinator would follow on a first call.

Logs live in your WordPress database with model name, token usage, and page URL captured per session. Most retirement planners export them to their CRM or to an archival system like Smarsh or Global Relay on a recurring schedule via webhook so chat lives in the same archive as email. You control retention, admin visibility, and whether summaries get sent to the planner after each session for the prep file.

Yes. The system prompt can ask about retirement timeline early in the conversation and route accordingly. Prospects 10 plus years out get pointed at pre-retirement content, prospects within 5 years get fast-tracked into a planning discovery call, and prospects already retired get routed to the income-planning track. Each path can use a different system prompt, a different scheduler, or a different set of follow-up emails.

Claude Sonnet and GPT-4o both follow the no-specific-recommendation prompt reliably and produce consistent disclosure language. They also handle the technical mechanics of Social Security, Medicare, and tax-bracket math more accurately than smaller models, which matters because the cost of a wrong answer in this category is a prospect walking away with bad information. SleekAI is bring-your-own-key for OpenAI, Anthropic, Google, or OpenRouter.

Yes. SleekAI supports multibot deployment with 28 plus display conditions, so a pre-retirement bot can run on the accumulation pages, an income-planning bot on the decumulation pages, and a separate bot on the Social Security guide. Each bot has its own system prompt, knowledge sources, and disclosures, so the framing matches what the visitor is actually researching at that moment in the journey.