The Manifesto for AI & Agile Alignment

We are uncovering better ways to align agile teams with AI and each other, and to help others do so.

AI & Agile Alignment Values

 eXponential markets
  over incremental profits

  Scientific experimentation
  over big design up front

  Continuous throughput accounting
  over projects and budgets

  Autonomy in alignment
  over command and control

  Learning-flow acceleration
  over transformation frameworks

  Ecosystems of mutual benefit
  over one-way value streams

while items on the bottom have value, we value those on top more.

AI & Agile Alignment Principles

  1. Directly Responsible Individuals and Agents: one for each business, design, and technical goal. DRIs and AI agents must work together daily to learn and adapt to each others’ constraints.
  2. Capture and store learning in AIs: as small, self-organizing, cross-functional teams work to open the current constraint, they train AI agents to meet future ones.
  3. You Aren’t Gonna Need It: teams and AI agents continuously prioritize their work relative to the current throughput constraint, measure progress by it only, and defer all other work.
  4. Align small groups of teams and agents to form value streams that continuously adapt their goals and workflows to changing market constraints and to each other.
  5. Model rewards on mutual benefit across teams and streams to minimize silos, wastes, politics, and missed opportunities, and to promote shared resources, services and learning flows.
  6. Continuously refactor value streams: reuse, recycle or reduce all the resources each stream produces until all contribute to throughput and none to waste.

 

  1. Design breadth-first: use AI to proactively uncover repeating patterns in and between markets and value streams. These inform designs we can only detail as we learn more.
  2. Build factories, not products: business, design, devops and AI are only in the right relationships when their work aligns to open their mutual working and learning constraints.
  3. Simplify and automate: simple, automated solutions cost less to install, align, and maintain than complex manual ones, yielding more broadly applicable business outcomes faster.
  4. Use and value experimentation: evaluate experimental costs against competitive risks to innovate products that meet and create new market opportunities.
  5. Open platforms and APIs: under-served market segments and spaces between segments hold opportunities for efficiency and collaboration that new business will develop for itself.
  6. Transform to embrace change: continuously amplify your organization’s intelligence to meet changing market conditions and opportunities for new growth.

We acknowledge the Agile Manifesto and Permaculture Principles on which our work is based, and pay respect to their originators and pioneers.

XSCALE.CO

For AI and Agile
Alignment at Scale

To align AI with humanity we first have to align the humans with each other.

XP/BDD/ToC do this for a small team, enabling it to focus on its goals autonomously.

But how do multiple autonomous teams align to form an autonomous value stream?

How can these streams continuously align to VUCA markets and each other?

How do AI agents integrate into these teams and streams to form a self-aligning organization?

We can’t afford Bureaucracy any more

Command & control committees multiply costs of change, quality, and delay by 10x.

Big meetings constrict learning flows and make majority-rule compromises, killing outcomes.

Doer-decider distance garbles and delays learning, making SNAFU and breaking autonomy.

Hierarchy slows lateral learning so the left hand doesn’t know what the right hand is doing.

KPIs and OKRs optimize focus on local metrics, not on growing end to end business throughput.

We need an Ecosystem,
Not a Framework.

A mutual-benefit network to align Business, DevOps and AI to growing end-to-end throughput.

Throughput-accounting focused on opening key constraints today, not “what you’re gonna need.”

Pull-based change working quicker, cheaper and safer – ie. less risky – than any framework.

Without pushing change into pre-existing teams, you don’t wait on resistors or compromise methods.

Instead, you invite green people into mature teams to learn how it works with peers, not “masters”.

Scroll to Top