A Tribre playbook · Installed in four weeks

Your team already uses AI. Your work was never designed for it.

A four-week operating model redesign for software companies of 10–60 people. Every recurring piece of work gets a best owner — a person, an AI, or the two working together.

Human + AI · one owner per piece of work

The problem

Nobody asks for work allocation.

Teams ask for relief from the symptoms. In a founder-led software company where everyone already uses ChatGPT, Claude, Cursor, and a growing set of agents, work isn't broken — it just outgrew the way it was divided. We hear the same sentences again and again.

“Everybody is building their own AI workflows.”

“We don't know what can be delegated.”

“Knowledge lives inside people's heads.”

“Every project starts from scratch.”

“People keep interrupting each other.”

“Nobody knows who should own new work.”

“Documentation is outdated.”

“We're hiring while AI could probably do half of this.”

The founders feel it first — they become the bottleneck. Not because work is broken. Because AI suddenly made delegation possible.

I want every recurring piece of work to have a best owner.

What the founder actually wants — one way or another.

That owner might be one person, several people, an AI, or a human working with one. The technology is beside the point. The outcome is less friction — and clarity about where work should live.

One personSeveral peopleAn AIA human + AI workflow

Who it's for

Small software companies, already deep in AI.

The playbook is deliberately narrow. It is not an organizational operating system for everyone — it is built for one environment where work allocation becomes valuable fast.

The company

  • Between 10 and 60 people, usually founder-led.
  • Already using AI extensively — but in an unstructured way.
  • Everyone has their own workflows. Nobody has the shared picture.
  • Founders increasingly feel they are becoming the bottleneck.

The shape

AI consultanciesSoftware studiosSaaS startupsDigital agenciesInternal innovation teams

People already work with ChatGPT, Claude, Cursor, Copilot — and increasingly with agents. What's missing isn't tooling. It's knowing who owns what, where work should go, and what can safely be delegated.

The playbook

An operating model redesign, installed in four weeks.

Not consulting. Not AI strategy. Six phases that move from how work actually happens today to reusable human + AI workflows — each phase producing something you keep.

01

Week 1

Discovery

Understand how work currently happens — through interviews and everything the company already has: SOPs, project boards, Slack, Notion, Linear, Google Drive.

InterviewsExisting SOPsProject boardsSlackNotionLinearGoogle Drive

Output → Current Work Map

02

Week 1–2

Work inventory

Every recurring activity becomes a Work Item. The company suddenly sees itself clearly.

“We actually perform 350 recurring activities.” That alone is valuable.

Review pull requestWrite proposalPrepare invoiceHandle support ticketRun sprint planningInterview candidateDeploy releaseApprove expenses+ 340 more

Output → Work Inventory

03

Week 2

Role extraction

Instead of looking at people, we look at work. Each Work Item belongs somewhere — and roles emerge naturally. Not John, Mary, and Lisa. Roles.

Product Discovery

Sprint Planning

Release Management

Customer Success

Finance Operations

Knowledge Steward

Prompt Engineering

Agent Maintenance

Output → Role Library

04

Week 3

AI assessment

Every Work Item is assessed against five questions. The answers become a delegation score — a grounded view of what AI should and shouldn't touch.

01

Can AI fully perform this?

02

Can AI prepare it?

03

Should AI review it?

04

Should AI only assist?

05

Should AI never perform it?

Output → delegation score per Work Item

05

Week 3

Work allocation

Ownership is redesigned. Instead of one person carrying everything, each role holds a collection of work — and each collection goes to its best actor.

RoleCollection of workBest actor
HumanAIHuman + AITeamExternal partnerFuture hire

06

Week 4

Build workflows

Now the interesting part. Allocated work becomes reusable workflows, with humans and AI each doing the part they're best at — like this support flow.

  1. Ticket receivedSystem
  2. AI classifiesAI
  3. Knowledge lookupAI
  4. Draft responseAI
  5. Human approvesHuman
  6. AI updates documentationAI
  7. Metrics updatedSystem

One reusable workflow

Deliverables

What you walk away with.

Seven artifacts — each one useful on its own, together a complete picture of how your company works and where the leverage is.

01

Work Inventory

Everything recurring inside the company, made visible.

02

Role Library

Every operational role, extracted from the work itself.

03

AI Opportunity Report

The highest-ROI automation opportunities, ranked.

04

Work Allocation Map

Exactly who — or what — should perform each piece of work.

05

Knowledge Map

Where knowledge lives — and where it is missing.

06

Agent Portfolio

Current agents, missing agents, and the ones to build next.

07

Implementation Roadmap

Quick wins30 days90 days6 months

Where Tribre fits

The playbook ends with artifacts. Tribre keeps them alive.

Not another documentation tool. Not another project board. Tribre becomes the living operating system — the source of truth for how work is allocated, and how that allocation keeps improving.

When roles change, when a new agent arrives, when a workflow stops working — the map changes with you.

  1. Work
  2. Roles
  3. Actors
  4. Knowledge
  5. Allocation
  6. Continuous improvement

Success metrics

Measure clarity, not AI adoption.

Adoption numbers say little about whether work got easier. Organizational clarity does — and it can be measured.

  1. 01

    Recurring work documented

  2. 02

    Work assigned to roles

  3. 03

    Work with clear ownership

  4. 04

    Work AI-assisted

  5. 05

    Work fully automated

  6. 06

    Time spent searching for information

  7. 07

    Time to onboard a new employee

  8. 08

    Recurring interruptions

  9. 09

    Undocumented workflows

  10. 10

    Duplicated activities

The first engagement

The AI-ready Work Allocation Sprint.

Deliberately small and outcome-focused. Two weeks, your team and ours, ending in an executive presentation of where work should live and what to change first.

2 weeks5–15 participantsFixed scopeExecutive presentation

You walk away with

  • A complete work inventory
  • 20–40 recurring workflows documented
  • 10–20 operational roles identified
  • AI delegation assessment for every workflow
  • An initial Work Allocation Map
  • Prioritized automation opportunities
Book the sprint

Tell us a little about how work happens today. We read every note.

Clarity about where work should live.

Know what can be delegated, what should stay human, and which changes create the greatest leverage.

Book the two-week sprint