How we work

Know what you're committing to, one stage at a time

Our value starts with the engagement. Every stage has a defined scope, a set price, and a deliverable you keep.

01 / 60 minutes · no charge

Discovery call

A working session, not a demo. We walk your current stack, the workflows that consume the most operations time, and where the handoffs break. You will know which workflows are worth automating and which are not — including the ones where the honest answer is to fix the process rather than automate it.

What you get
Our read on where the time is going.

02 / Two weeks · fixed fee

Workflow consulting assessment

We sit down with your team and dig into the manual work that costs your firm the most. You get a written report that lays out where the hours are going, which of that work we can automate, what the payoff would be, and a step-by-step plan to get there.

What you get
The report, to keep and act on regardless of what you decide about Leverance.

03 / 6–8 weeks · monthly fee

First agent live

We set up your agent, connect it to your systems, build in your approval rules, and run it with your team watching, all before it touches a single client record. Nothing goes live until you have seen it work and said yes. From then it is the platform fee plus the agent fee, monthly.

What you get
A working agent, an audit trail, and a documented rollback.

Access

What access we need, and when

Access widens only as the engagement progresses, and never beyond what the workflow in front of us requires.

StageAccess
DiscoveryRead-only, or a screen share. No credentials change hands.
AssessmentRead-only, and only to the systems that workflow actually uses.
BuildCredentials for the systems the agent touches, with only the permissions that workflow needs.
LivePermission to write only what the workflow writes, and nothing beyond it.

Questions

The questions we get asked first

Roughly in the order they come up.

Is our client data used to train your models?

No. We never use your client data to train or improve any AI model, and your firm's data is never mixed with another firm's. The AI model runs inside your firm's own cloud subscription, provided through your cloud's marketplace under the model provider's own terms, which your firm accepts directly; your firm can arrange zero data retention with that provider and confirm it with them. Leverance never holds the credentials for the model. Before anything goes live, we give your compliance team a written schedule of what the model sees, what Leverance can see, and every provider involved in supporting you, so it is on file rather than a verbal assurance.

What happens when the agent gets something wrong?

It depends on the kind of mistake, and there's a safeguard for each. Obvious errors, like a value in the wrong format or a required field left blank, are caught by automatic checks before anything is submitted. The agent stops and hands the item back to a person rather than letting a custodian reject it later. Subtler mistakes, where a value looks reasonable but isn't right, are what the approval step is for: an advisor reviews the proposed action, and where the information came from, before it goes through. Either way, nothing reaches a client record without a person signing off.

Do we have to disclose this to clients, or in our Form ADV?

That's a question for your compliance counsel and depends on how significant the AI's role is in the advice you give. What we can do is make the disclosure easy to write accurately. We give you a plain-language description of each agent: what it does, what it doesn't do, and where a human makes the decision. That matters because the SEC's enforcement actions on AI so far have targeted firms that overstated what their technology does. Describing it exactly is safer than describing it well.

How much of our team's time does this actually take?

Most of it comes at two points, and both need knowledge only your people have: the initial assessment, and deciding who approves what. For those, we'll need your operations lead and whoever does the work today, in a handful of scheduled sessions rather than an ongoing commitment. Connecting the agent to your systems is our job, not yours. And once it's running, it shouldn't need regular attention from anyone senior; if it does, we've built the wrong thing.

What do we keep if we stop working with you?

Everything except us. The full record of what the agents did stays with your firm, which matters because your books-and-records obligations don't end when our engagement does. The assessment report is yours from the day we deliver it, whether or not you go further. And how we hand things off is spelled out in the agreement before you sign, not worked out at the point you decide to leave.

What happens if Leverance goes away?

You keep running. The software and the agents are installed inside your own environment, not on our servers, and your client data never leaves it, so nothing of yours disappears with us. The work the agents have done is already in your systems, along with the complete record of what they did. The agents keep working, too: if Leverance ceases operations, your license becomes permanent for the version you have installed, with no renewal that depends on us. You would stop receiving updates and support, but the software keeps running as is. These terms are spelled out in our Services Agreement, and we're happy to walk you and your counsel through them.

Let's talk

We love talking about RIA operational problem areas and how to fix them. Put our team to the test.

Book a 60-minute discovery call