We design, implement, and support AI agents, AI assistants, and automated workflows that unlock capacity and capabilities for the mission-critical people and systems that power your practice.
Small and mid-sized professional services firms: practices where the work product is judgment, and where necessary production work occupies the people who provide it.
Across the landscape of practices: estate planning, immigration, family, business, real estate, litigation. The details differ by practice area; the shape of the day rarely does. Intake, deadlines, document preparation, and client questions fill the hours the legal thinking should get.
Solo practitioners through multi-partner firms. The work arrives in seasons. Client documents, preparation, and review stack against fixed deadlines, and the advisory work that deepens a client relationship waits for a quieter week that rarely comes.
Independent advisory and wealth management practices. Client reviews, account paperwork, and the platforms behind every account fill the calendar, and the conversations that actually move a client's plan compete for what is left.
Vendors build the capabilities they can sell across 80% of the market. What remains is a gap in their roadmap for solutions specific to how your practice operates. humanagentic builds that missing 20%, unlocking capacity with agentic solutions and enabling human judgment to deploy where it's needed.
No two practices are the same, so our process begins with a conversation about your practice and how it actually runs, understanding the problem space, identifying the pain points, and aligning on the problems to solve. Below is an example of what that conversation might sound like.
Honestly, the part that gets me is the review packets. My team assembles each one: performance from Orion, the plan from the planning software, the notes from the CRM, then checks it all against Schwab. And I read every one before it goes to the client.
That assembling takes a trained person, but none of it is advice. Your read is where the judgment goes. What's that costing you at the other end?
That's right. And the same two people build the packets and do the onboarding. So we've stopped taking referrals. As of today, three families essentially waitlisted, because onboarding one runs weeks and every form the custodian sends back costs days. And we priced the platform everyone says to buy. Per seat, annual, a minimum bigger than our whole firm.
If I may play that back, and fix anything I get wrong. You're not short of demand, you're short of capacity to take it. The assembling can come off your people, and their week goes to the onboarding. The reading stays with you.
Let me back up, I don't think I said what the read is for. I'm looking for signals a family's situation has changed. We send a questionnaire ahead of every review, and it mostly doesn't come back. The change comes out in the meeting, after the packet's built.
Got it. So what changed decides what the family needs, and it's the last thing you learn. That's what's worth fixing, and it isn't the packets.
You're right, it isn't. So what would fixing it look like? Because I've been sold this before.
Fair. I'm hearing two problems: taking on the families who are waiting, and learning what's changed for the ones you already have. And I won't pitch you either fix today. What the fixes look like is what discovery is for: we sit down with your team and map the workflows, the systems, how you handle risk, and where your judgment goes. And regardless of what discovery finds, how we build doesn't change: the agents carry the workflows, deciding what a family needs stays yours, and nothing reaches a client without a human approving it, every step recorded. How does that sound?
That's the conversation I want to have.
01
Where is high-value judgment pinned under necessary work in your practice, and what would it take to unblock it. We map the workflows, the systems, and the people they occupy.
02
What to build, in what order, and why. The thesis applied to the unique way your practice operates: a sequenced plan your team can absorb, with the consequential decisions and their owners identified up front.
03
We design, build, and deploy the AI agents, AI assistants, and automated workflows the strategy calls for, working with the systems and people already in place.
04
We keep the systems current, monitored, and improving as your practice changes. New needs feed back into strategy; the relationship continues.
Your firm is accountable for delegated work whether it goes to a new associate or to software. So the question is not whether AI does the work. It is where your people stay in the loop, and what is escalated to them. We build to guardrails you establish and practice standards you follow.
The steps are set with your team before anything runs, and nothing happens outside them. Models read, sort, and draft within a step, and can choose among options your process already allows; they cannot add one. The same input is handled the same way every time. Where your process calls for a person to approve, it stops and waits.
The agent decides its own next step, works until the job is done or a limit stops it, and corrects itself when a step fails. The limits are written first: what it may do alone, what requires approval, what it may never touch. Consequential decisions are escalated to a named human, with the record of how it got there.
It draws only on materials you have approved, cites the source of every answer, and states when it does not have one. It sends nothing, files nothing, commits nothing. A human decides, every time.
Each is built as a workflow, an agent, or an assistant. Which one depends on how your practice runs.
Not an exhaustive list.
Jerry has spent over fifteen years building and growing software applications and products for some of the most recognizable names in media: Gannett-USA Today, AddThis, Politico, Bethesda Softworks, Vox Media, and Yahoo-AOL.
At Vox Media he built the subscriber acquisition and retention journeys behind NYmag.com, a portfolio of products and features that generated $15.6M in annual recurring revenue; he also implemented the AI that powers its conversion funnel, dynamically deciding which offer to show a prospect based on their onsite behavior. At Yahoo-AOL he built a cross-device commenting and community platform that turns readers into participants, earning $936K in advertising revenue. At Politico he built the first-party audience data applications behind $3M a year in advertising, and developed the business and investment strategy for the site's paywall. At AddThis he rebuilt the publisher tools platform, the monetization engine for a business that served fifteen million websites and held the largest data footprint on the internet behind only Google and Facebook (Meta); the rebuild lifted product activations 140%. Earlier work includes a streaming and audiobook platform delivered to more than five thousand libraries at RBmedia, recommendation and loyalty systems at Bethesda Softworks, and the products that grew USA Today's Travel and Sports sections past one hundred million monthly users.
Each of those results came from adopting new technology early and putting it to work: personalization and recommendation algorithms before they were table stakes, first-party data platforms before the industry followed, conversion rate optimization programs based on strategic planning and statistical discipline. He now builds with agentic AI, deep in the current generation of models and tooling, delivering complete, production-grade applications, taken from specification to launch and serving real users. The result is the strategic thinking, data-driven discipline, and process-driven execution of an enterprise-level agentic builder, applied to your practice.
humanagentic exists because agentic development changes the economics of custom software. The software a firm actually needs is shaped by how that firm works, and building it that way has never been economical for vendors, which is why the market offers generic products that fit no one or enterprise engagements priced for someone else. That arithmetic has changed. Custom, enterprise-level solutions are now economically viable for small and mid-sized practices, and building them is what this company does.
Jerry holds an MBA from Nova Southeastern University and a B.S. from the University of Florida. He lives in the Washington, D.C. area with his family.
Over fifteen years building software for some of the most recognizable names in media: Gannett-USA Today, Politico, Vox Media, and Yahoo-AOL. Jerry built custom systems that generated new customers and increased the value of existing ones. He founded humanagentic to bring the rigor, processes, and innovation of enterprise-grade software development to professional services, to enable business growth and deeper client relationships.
This part of the page needs JavaScript to run. Turn it on and reload, and the questions will be here.
As your role at your practice, here is what you want off your plate: the work that doesn't need your judgment. Why it matters: what that time would go to. What is in the way: the thing stopping you.
