AI fluency for ambitious teams

You don't need another AI tool. You need a team that can drive one.

AI has come far enough. The tools you already pay for can do the work. What's missing is a team that knows how to use them, and that's a learning problem, not a shopping problem. We audit your business, train your people, and only talk about building things once the fluency is real.

Apply to work with us See how we work

Works where your team works:Microsoft 365Google WorkspaceSalesforceHubSpotMonday.comSlackQuickBooksNetSuiteYardiDocuSignNotion

The problem

The models kept getting better. Most companies didn't.

Here's what we keep finding inside good companies. Licenses that mostly go unused. One person in ops quietly doing the work of three because she figured it out on her own. A pilot from last year that never left the demo. And a calendar full of vendors pitching the next platform.

Nobody in that building needs new software. The AI they have can already read the contracts, draft the quotes, reconcile the statements, and write the code. Fluency is built, not bought, and right now it lives with whoever happened to figure it out alone.

Our whole business is closing that gap, department by department, until the capability belongs to you and not to a vendor.

Step 1

The audit: two weeks, mostly conversations.

We sit with leadership, then with a few people in each department, and map how work actually moves. Not a questionnaire. Conversations.

You get an honest read on where your company stands with AI, where the hours and dollars leak, and a ranked list of what fluency would be worth, in dollars. Fixed fee. The findings are yours either way.

Fig. 1 — where the hours go

Step 2

Then we make your team good.

Live trainings on their real work. Hackathons where departments race to automate their own bottlenecks, with our people in the room.

Alongside, we build skill files: living playbooks that teach the AI how your company writes a quote, closes the books, answers an RFQ. What your best people know stops walking out the door.

quote-builder.md month-end-close.md client-onboarding.md

skills/quote-builder.mdv3 · live

when: a customer asks for pricing

pull: price sheet, freight rules, terms

format: the way Maria writes them

never: quote below floor without approval

# written with the sales team, week 2

Fig. 2 — a skill file, in plain language

Step 3

Then, and only then, custom builds.

Once the fluency is real, we look at what's worth building. By then your team can tell the difference between a tool they need and a tool they'd abandon, and the build lands on people who will actually run it.

Most firms sell you the build first. We think that's backwards.

team trained on live workflowsDONE

skill files in daily useDONE

hackathon winners in productionDONE

custom build: client intake agentNOW WORTH IT

Fig. 3 — builds wait for fluency

What you get

Three things you can point to.

.01

You'll know your level

Every department measured against what AI can actually do today. No vendor optimism, no doom. A number and a reason.

.02

ROI before spend

A ranked list of opportunities in dollars, built from talking to your people, before you commit to anything.

.03

Skills that stay

Trained people and skill files that keep working after we leave. The capability is yours, not a subscription.

The stack

Claude first. Fluent in the rest.

Claude & Claude Code

Our home base, as part of the Anthropic Partner Network. Claude for daily work. Claude Code and the CLI for the people who build. MCP to connect it to your systems. Skill files, projects, agents, governance.

ChatGPT

Fluent. Enterprise rollouts, custom GPTs, and team training where your company already runs on it.

Microsoft Copilot

Fluent. Most companies already pay for it and use a tenth of it. We make the license you bought earn its keep inside Excel, Outlook, and Teams.

The rest of your stack

Perplexity, automation tools, whatever your work touches. We do not sell software and we do not take referral fees. The recommendation is just the recommendation.

Who it's for

Companies that make and manage real things.

.01

Real estate & financial services

Owners, operators, lenders, funds, and managers. Underwriting support, reporting, diligence, client service, back office.

.02

PE-owned companies

Value creation without new headcount. We work with operating partners and management, and the playbook travels across the portfolio.

.03

Manufacturing & consumer brands

Product companies with real operations: procurement, inventory, wholesale, retail compliance. The teams most AI vendors never bother to visit.

.04

Healthcare operations

Administrative and operational workflows, handled de-identified and governance-first from day one.

“The knowledge of how your business runs is already in your people's heads. Our job is to connect it to the AI you're already paying for.”

Daniel Gofman · Founder, The Implementation Squad

FAQ

Fair questions.

What actually happens during the audit?+

Mostly conversations. We start with leadership, then sit with a few people in each department and watch how the work really moves, which is usually different from how anyone thinks it moves. Two weeks later you get the readout: your level, your leaks, and a ranked list of what's worth doing, in dollars.

Do we need to buy new software?+

Almost never, and that's sort of the point. Most companies are sitting on more AI capability than they use. If something in your stack should change, the audit will say so, but we don't sell software and we don't take referral fees, so the recommendation is just the recommendation.

What's a hackathon like?+

A day or two where your departments compete to automate their own bottlenecks, with our squad in the room. Nobody sits through slides. It's loud, it's competitive, and the winning work usually goes into production the same week. It's also how skeptics convert, because it's their own workflow that got faster.

What are skill files?+

Plain-language playbooks that teach the AI how your company does things: how you write a quote, close the month, answer an RFQ, handle a complaint. Your team writes them with us during training. They live in your systems, they improve with use, and they keep working long after we're gone.

We already ran an AI training. It didn't stick.+

A lunch-and-learn isn't training. We train people on their own work, test them on it, and come back on a schedule to re-measure. Skills stick when they're used weekly and someone checks. That's the program.

When do you build custom things?+

After fluency, not before. A custom build dropped on an untrained team becomes shelfware in a quarter. Once your people are fluent, they'll tell you exactly what's worth building, and they'll be right. Then we build it with them.

How does pricing work?+

The audit is a fixed fee, agreed before we start. Training and hackathons are scoped by cohort. Builds are scoped by project. Nothing open-ended, no surprise invoices.

Why do I apply instead of booking a call?+

Because this only works when both sides commit. Changing how people work takes real effort from your team, and we cap how many engagements we run at once. The application is five questions and takes about two minutes. If we're not the right fit, we'll say so and point you somewhere useful.

Work with us

Changing how people work is hard. We only take on teams that want it.

Your people already have a full workload. Now they're being asked to learn new tools and unlearn habits that took years to build. That only works when both sides commit. We're hands-on, we cap the number of engagements we run at once, and we treat every one as a partnership.

So we don't do a sales call. Tell us where your company stands and what you want your people to be able to do. We'll tell you honestly whether we can get you there. Replies within two days.

Apply to work with us

Five questions · two minutes · every engagement starts with the audit

Or write directly: daniel@theimplementationsquad.com