Index 03 — Approach

Three steps. 10 to 80 days.

A network firm runs differently from a consultancy and differently from a marketplace. Here is exactly how a clus.ai engagement works, from the first email to the day we hand off.

i. Scope

Two weeks. Fixed fee. A senior partner runs it personally.

The diagnostic identifies which workflows in your business are agent-ready today, which need data or process work first, and which should not be touched yet. Most firms charge six figures and three months for this work. Ours is concrete, scoped, and ends with a written report your team can act on whether or not you continue with us.

You leave the scope phase with three things: a prioritized list of agent-ready workflows ranked by expected impact, a candid map of the data and process work required for the harder ones, and a fixed-price proposal for the deployment phase if you want one.

If a workflow isn't ready, we tell you. The fastest way to lose a year is to deploy an agent on top of a broken process.

ii. Match

From an invitation-only network of senior operators, we assemble a small pod.

Pods are matched to your industry, your stack, and your stage. A revenue agent for a Series C SaaS company is a different pod from an M&A integration for a PE-backed industrial roll-up. We don't have a bench, which means you don't pay for one. Operators come in, ship the work, and rotate out. The senior partner who ran your scope stays on as the accountable lead.

The network is being built in public. Every operator we add has shipped agents in production at enterprise scale. We turn down more applicants than we accept, and we'd rather move slowly with a small senior network than fast with a big mediocre one.

iii. Ship

The pod works in production from week one. Embedded 10 to 80 days.

The deliverable is an agent your team owns. Not a deck. Not a roadmap. A working system with documented prompts, eval suites, monitoring, and the architectural choices written down so your team can extend it.

We end the engagement when the target metric moves and your team can run the system without us. Outcome-tied. The cleanest measure of whether we did our job is whether you stop needing us.

What "10 to 80 days" actually means

The range is wide on purpose. A focused single-workflow agent on a clean stack ships in 10 to 20 days. A multi-system deployment with permissions modeling, data plumbing, and change management runs 60 to 80. Most engagements land in the 30 to 50 day band.

We will not stretch an engagement to fill a calendar. The pod ends when the work ends.

What we don't do

  • We don't sell strategy decks. The scope phase produces a written assessment, not a 60-slide deliverable.
  • We don't bill by the hour. Engagements are fixed-fee, with a performance component tied to the agreed metric.
  • We don't lock you to a platform. Model-agnostic, vendor-neutral, your IP from day one.
  • We don't carry a bench. When operators aren't on engagements, they're operating. That's why they stay sharp.

Pricing

Scope: fixed fee, two weeks. Quoted on the call.

Deployment: fixed fee plus a performance component tied to the scope metric. We share the upside when the metric moves, which keeps incentives honest on both sides.

Operator placement (retained search for AI-native talent): standard retained search structure.

Have a workflow worth automating?

Reply within 48 hours

Tell us what you're trying to ship. A senior partner replies within 48 hours. No SDR, no sequence.

Request a brief

San Francisco · New York · London · Paris