01 · work theory

Most studios design what you see.
We design how you work.

Work Theory is an AI enablement and infrastructure practice for startups, scaleups, and SMBs. We put AI inside your team's real workflows, build the rails underneath, and redesign how your org learns, so the capability compounds after we leave. Agency, not dependency.

02 · thesis

Adoption is no longer the problem. Scaling is. Pilots are everywhere. Practice is almost nowhere.

01 the scaling gap

88%

use AI somewhere. Only a third are scaling it.

Most "AI-enabled" orgs are running isolated pilots, not transformed operations. Two-thirds haven't started scaling at all.

mckinsey · state of ai 2025

02 no measurable impact

95%

of orgs see zero profit impact from AI investment.

An NBER survey of 6,000 global executives found 89% see no effect on labor productivity. The technology is real. The translation layer is missing.

mit · gallup · nber, q1 2026

03 redesign beats access

2.8×

High performers are more likely to redesign workflows, not buy more tools.

They're also 3× more likely to have leaders actively championing AI use. The differentiator is how the org reorganizes around what the tools can now do.

mckinsey · ai trust maturity, mar 2026

04 managers are load-bearing

12%

of employees say AI has actually transformed how work gets done.

Manager-led adoption is a top driver of frequent AI use. The gap between deployment and durable practice runs through the middle of the org chart.

gallup workforce, q1 2026

<summarized by claude · mckinsey, anthropic economic index, gallup, nber, mit>

The orgs that close this gap won't do it with a course library. They'll do it by redesigning how their people learn, inside the work, with peers, at the speed the tools change. That's the theory we work on.

Underneath it, a simpler belief: when work is an expression of who you are, it becomes what you want it to be. We build teams that get to work that way.

03 · services

Enablement on top. Infrastructure underneath.

<track 01 · workflow enablement>

For teams putting AI inside the actual work.

We start with the work, not the tool. Your team brings real tasks; we build the AI-handled version together, then turn it into shared practice with rituals, reviews, and a cadence managers can run.

Engagements start with a four-week pilot on a single workflow your lead can defend, and scale with proof: a full unit, a department, the org. Your people own the work after we leave.

scope & cadence matched to team size and rollout footprint

<track 02 · ai infrastructure>

For orgs whose adoption has outgrown the chat window.

Durable practice needs rails. We design and build the layer your team works on: internal agents, prompt and context libraries, integrations into the systems where the work already lives.

Built with your team, not for them. Every piece ships with the documentation and training to run it without us. Infrastructure your people can extend is infrastructure that lasts.

built in your stack · owned by your team

<track 03 · product & post-sales enablement>

For teams that bought the tool but can't staff the rollout.

You signed the contract. The vendor handed you a quickstart and a Slack channel. Adoption stalls at the team you didn't have time to train.

We come in as the adoption layer between vendor and end user. Embedded enablement, custom workflows, manager rituals, and a usage feedback loop the vendor can act on. No new headcount.

scope matched to contract size and rollout footprint

04 · who it's for

Startups, scaleups, and SMBs.

Startups & scaleups

<no L&D function · adoption is uneven>

Pockets of power users; everyone else watching. We embed alongside the team, get AI into the workflows that matter this quarter, and close the gap with shared rituals, manager-led cadence, and infrastructure that makes the good patterns the default. Practice, not process debt.

SMBs

<lean team, every hat>

Owner-operators with no ops layer and no time to build one. We find the few workflows that buy back real hours and make them stick, so AI is this quarter's leverage, not next year's project.

05 · team

Two practitioners. No middle layer.

Daye

<partner>

Learning systems engineer. Previously Slack, Salesforce, Uber, Vonage. Across GTM, Support, Ops, People, Product. A lifelong supporter of animal welfare, from a heritage of artists, educators, and entrepreneurs.

Nelson

<partner>

Learning systems engineer. Previously Meta, TikTok, Amazon, Slack. Across GTM, Marketing, Ops, Legal. A lifelong educator who loves being in front of a classroom, and is learning Portuguese just because.

Start with a working theory.
Finish with a working team.

We take on a small number of engagements per quarter. Tell us the workflow you want to change and we'll send back a scope.

or email hello@worktheory.ai

Optional, and it stays between us — no newsletter, no list. We read every one and reply ourselves.