my first williamfrank agent is live.
I'm an investor, not an engineer. I didn't write the code — WilliamFrank, the agent I built, did: the site, the design, even its own backend. I described what I wanted. It built.
This is the proving ground. Next, an integrated WilliamFrank across my ventures. Link below.
The best AI model changes every few months.
If you run a wealth team, each change can mean retraining your people on a new tool.
Own the record of each family and write your method down.
Then agents can use whichever model is best this quarter, and your team keeps one way of working.
That is how a small firm keeps up with the models without starting its team over each time.
(Drafted with AI, edited and refined by Bill Woodruff)
For the owner of a small RIA, AI governance adds a review to an already full week.
An agent that records what it did, and who approved it, shortens that review.
Firms that build the record in from the start should spend less of the exam explaining. https://t.co/hYmrM1eCug
The SEC's exam division plans to probe how firms govern AI use, from deepfake threats to vendor risk, while also prioritizing illiquid products and Reg S-P readiness. https://t.co/eEW9ZypmEA
Many RIAs are a handful of people, with the owner running the systems and managing the team.
That owner has the most to gain from an agent, and a 20-user minimum leaves them out.
Curious whether Schwab revisits it before IMPACT. https://t.co/e34i3qKNNJ
Schwab trumpeted Anthropic's Claude for all 16,000 of its RIAs, but a 20-seat enterprise minimum suddenly sidelines 11,000-plus small firms -- right before IMPACT 2026 https://t.co/H9bZpNEJyF
AI gives you more of whatever you bring to it.
Use it as a shortcut and you get faster at being wrong. Use it to build and learn and you get better at the work.
From Episode 2 of the WilliamFrank podcast. Full episode in the reply.
A wealth firm's operations are mostly checks.
Did the transfer settle? Did the signed form come back? Is the allocation still inside its bands?
A person runs each one on a schedule.
The leader checks that they did.
Agents can run them all the time, on a record that shows when an answer changed.
That leaves the leader's week for clients and the decisions no check can settle.
(Drafted with AI, edited and refined by Bill Woodruff)
@LexSokolin For the person running the firm, the license is the smaller part of the bill.
The larger part is the time spent managing the people who keep the systems in step.
An independent vendor that takes some of that off the owner's week has an answer worth paying for.
@AaronKlein In a wealth firm, most of the scheduling is the client review cycle.
An agent that books those meetings is one less thing to keep track of.
Reschedules can be tricky with back and forth.
@TDBank_US's U.S. wealth unit hired @hardeepw, Schwab's former head of firmwide AI, to run the business.
A bank picked the person who designs the work to run the firm.
Whoever runs a wealth team already holds both jobs.
Firms that design the work around agents will give their leaders more time with clients.
(Drafted with AI, edited and refined by Bill Woodruff)
@TDBank_US's U.S. wealth unit hired @hardeepw, Schwab's former head of firmwide AI, to run the business.
A bank picked the person who designs the work to run the firm.
Whoever runs a wealth team already holds both jobs.
Firms that design the work around agents will give their leaders more time with clients.
(Drafted with AI, edited and refined by Bill Woodruff)
In a wealth firm, the bottleneck is often the owner's calendar.
Every client decision and every review runs through it.
Agentic wealth helps when it brings those decisions forward, ready to make. https://t.co/uaeFwssycm
This is an interesting question because the answer is not obvious.
Which, by the way, does not mean AI coding tools are not leading to massive improvements, but, as research keeps showing, the bottlenecks to product improvements are often elsewhere.
A $20B firm can build its own agents and staff a team to run them.
At most smaller RIAs, the owner would be that team.
What they need is the agents without a new job for the owner. https://t.co/BXoUVxfYpG
If you run a wealth team, most of the AI on offer connects to the systems you already have.
Ask it for meeting prep and it finds a client's move.
That is usually the day before the review.
The updates still go to four desks, and checking them falls to you.
A harness catches the move the day it arrives and drafts the updates.
That leaves your week for the decisions.
(Drafted with AI, edited and refined by Bill Woodruff)
@LexSokolin In a wealth firm, the hard part is that advisor being free when the call comes.
Their week fills with tracking and checking on the team.
Agentic wealth leaves that time for the call.
@AaronKlein At a small wealth firm, most hires get screened for follow-through on tracking.
If agents carry the tracking, an owner can hire for judgment with clients.
Are you seeing a change in how you think about hiring since Riskalyze?
Where did the work go once AI did the building?
Into deciding. A prototype like this is a few hundred decisions, and most of them took longer to make than they took to build.
From Episode 2 of the WilliamFrank podcast. Full episode in the reply.