Chicago, wealthtech product launch, announced September 10.

FinTurk, an AI-first CRM built for registered investment advisors, launched two new features this week: PortfolioSolver, which turns an advisor’s plain-language client instructions into a proposed set of trades, and Vigil, a background monitoring layer that flags stalled workflows, failed data syncs and unanswered client emails before they become a problem. Both build on a product FinTurk launched only in May, when it debuted with Chicago Partners, an $8 billion RIA, as a founding client.

The pitch with PortfolioSolver is that an advisor can type something like “keep realized gains under $10,000 this year and don’t touch the Apple position,” and the system translates that into structured rules, runs a deterministic optimizer against them, and produces a trade proposal with a stated rationale attached, all before the advisor manually builds anything in a separate portfolio management tool.

Vigil is the more mundane but arguably more useful of the two: it’s a monitoring layer that checks connections, portfolios, tasks and inboxes on a recurring basis and surfaces what needs attention, which is the kind of unglamorous plumbing that actually saves an advisor time day to day, compared to a flashier AI feature that gets used once a quarter.

FinTurk’s target market matters here. Independent RIAs are underserved by the incumbent CRM options, most of which (Redtail, Wealthbox, Salesforce Financial Services Cloud) were built as systems of record rather than systems that take action. FinTurk’s bet is that a smaller, faster-moving vendor can out-execute the incumbents on actually automating advisor workflows rather than just logging them, which is a real gap in the market, even if “AI-first CRM for advisors” is becoming a crowded pitch among wealthtech vendors this year.

The part that deserves real scrutiny is PortfolioSolver’s core promise: turning natural-language instructions into trades. Getting the interpretation layer right, correctly parsing what an advisor actually means by a loosely worded constraint, is exactly the kind of problem where a small number of edge cases (a client with concentrated stock positions, a wash-sale rule interacting badly with a stated instruction) can turn a time-saving tool into a compliance liability fast.

FinTurk says a human approves the translated rules before the optimizer runs, which is the right design choice, but it also means the tool is only as good as how carefully advisors actually read that intermediate step rather than rubber-stamping it, and rubber-stamping is exactly the failure mode advisors are trying to escape by adopting time-saving software in the first place.

FinTurk is moving fast, three product launches since May, which is either a sign of real momentum with its first big client or a sign it’s still figuring out what the product should be before it has enough scaled usage to know for sure. Both readings are plausible at this stage of a company’s life.

How PortfolioSolver Turns Plain-Language Instructions Into Trades

PortfolioSolver’s core mechanic is translating a loosely worded instruction, something like keeping realized gains under a set dollar figure while leaving a specific position untouched, into a structured rule set that a deterministic optimizer can act on. An advisor still has to approve the translated rules before any trade proposal goes out, which is the right design choice given how easily a misread constraint could turn into a compliance problem.

FinTurk has more on both PortfolioSolver and Vigil on its product site, and Fintechbits covers similar advisor-technology launches in its AI in finance section. Fintechbits covered a related open banking story today on Lumin Digital’s new MX integration, another example of vendors racing to modernize infrastructure ahead of regulatory deadlines.

Whether Chicago Partners or another named client is running PortfolioSolver in production, not just in a demo, by year-end is the thing that will actually tell you if this works.