AI's wealth payoff now runs through the client meeting

Practifi's chief executive has a word for the wealth management CRM: a taker, because for decades, as Adrian Johnstone argues, the system has demanded constant input and updates—the advisor gives, the software stores. This week his Chicago-based firm launched Sentir, an AI-driven CRM built to reverse that flow by briefing advisors on account status, preparing meeting materials, recommending next steps, and dragging in context from outside the system, as WealthManagement.com's WealthStack roundup reported. 'Advisors can view their CRM as a giver,' Johnstone said.

The launch is also a rebrand: Practifi's existing system becomes Practifi Naya and remains available to current and new users, while the firm says wealth managers can convert to Sentir 'with no friction.' Sentir runs 16 AI agents that analyze tax, financial planning, and client sentiment, adds a notetaker that captures video and in-person meetings, a meeting prep workspace that drafts notes to firm templates, and a timeline that consolidates every meeting, email, service request, and life event into a single view. Advisors can query the system by chat and connect it to other data sources.

The pitch marks the next stage of a shift already underway: the first wave of AI in wealth management lived in the back office, and this one wants to sit in the room. LPL's Latitude platform and Altruist's Hazel are already pushing agentic AI into the advisor's workflow, and Practifi is aiming at the meeting itself. But a giver only gives what the foundation allows. Client files at a typical RIA live in four systems, and AI needs them clean and in one place, which is why the data-warehousing question moved to the top of the stack. A notetaker that captures a meeting is useful only if the system also knows the tax situation, the financial plan, and the service history that made the meeting matter.

The same logic runs through the week's other stack news: Morningstar and PitchBook have signed a deal with Google Cloud to launch Gemini Enterprise for Financial Services, which will fold source-attributed research and private market intelligence into Google's enterprise AI. The integration, available soon, gives subscribers Morningstar's public-market investment data and research alongside PitchBook's private-market data on companies, investors, funds, and transactions, and Seth Sprinkle, Morningstar's global head of AI platforms strategy and partnerships, made the argument explicit—AI's value is determined by the information it draws on.

That is a distribution play dressed as an AI launch. Morningstar and PitchBook are routing their research and data through Google's enterprise sales machine, putting their content in front of Gemini users without having to build an advisor interface, and betting that the private-markets gateway is won at the data layer well before anyone opens a chat window. The same instinct drives AssetMark's newest feature, Talk Tracks, which converts portfolio data into client discussion; AssetMark reported $91.8 billion in regulatory AUM as of Aug. 22. Talk Tracks is a small feature with a large ambition—turning the portfolio review into a conversation the advisor does not have to script.

WealthReach and VastAdvisor extend the logic to prospecting, converting user search behavior into marketing outreach. VastAdvisor's insider SAFE made the case last week that distribution beats model quality, and this deal is the follow-through.

Across the week, the target has shifted from systems to the conversation itself: Practifi wants to brief the advisor-client meeting, AssetMark wants to script it, WealthReach and VastAdvisor want to fill it, and Morningstar with Google wants to supply the underlying facts. But none of these tools works on a messy book, so the RIAs that win will be the ones that did the unglamorous work of getting their data into one place before they let the agents loose. They will find out soon enough whether Sentir's 16 agents and Gemini's source-attributed answers are worth the price of entry.