---
title: "How to Scale AI Adoption with a Federated Paved Path"
description: "A five-step how-to for scaling AI adoption across engineering teams: federate ownership, ship one shared paved path, let builders own their tools, measure throughput, and feed bottlenecks back into the path."
---

# How to Scale AI Adoption with a Federated Paved Path

**Type:** [How to](https://yomu.fyi/visual-explainers/category/how-to.md)

![A five-step left-to-right flow diagram showing federate ownership, ship a shared paved path, assign builder ownership, measure throughput, and feed bottlenecks back, with a return arrow and a guardrail strip below.](https://yomufyi.s3.us-east-1.amazonaws.com/visual-explainers/50a04db1-3802-4f48-ae62-9677eb739f6f/image-240e6746-e424-459f-88e0-c037cf1d4372.webp)

> A five-step how-to for scaling AI adoption across engineering teams: federate ownership, ship one shared paved path, let builders own their tools, measure throughput, and feed bottlenecks back into the path.

## Explanation

AI adoption stalls when every team experiments alone, rebuilding the same prompts and duplicating effort. This explainer lays out a federated paved-path model: a steering committee sets direction, champions spread practice, and one shared plugin distributes instructions and skills across dozens of teams. Each capability is owned by the team that built it, and metrics connect emerging bottlenecks back to available tools. It also covers the guardrails that keep speed safe, including layered evaluation where different models review each other's work, human oversight, and the habit of reviewing agent output before accepting it.

## FAQ

### Why federate AI adoption instead of mandating one tool from the top?

A single mandate rarely survives contact with 70-80 scrum teams that each have different delivery commitments. The federated model keeps a steering committee and champions for direction, but lets teams contribute capabilities by solving their own pain points, with ownership assigned to the building team for end-to-end delivery and support.

### How do you keep AI-generated changes trustworthy as adoption scales?

Layer the evaluation rather than trusting one model to grade itself. Salesforce's Agentforce team used an agent-as-a-judge setup with different models reviewing work from different perspectives, plus human oversight, to avoid self-preference bias. Developer surveys point the same way: most Go developers using coding agents weekly still review AI-generated code before accepting it.

### Which metrics show AI adoption is actually working?

Track throughput, inner-loop efficiency, and cycle time rather than counting seats or prompts. In the Salesforce rollout, adoption approached 100% within three months alongside a 26% throughput gain, 32% better inner-loop efficiency, and cycle time cut from 22 days to 11 days. The feedback loop then converts new bottlenecks into the next paved-path capability.

**Tags:** [Adoption](https://yomu.fyi/topic/adoption.md), [AI Adoption](https://yomu.fyi/topic/ai-adoption.md), [Change Management](https://yomu.fyi/topic/change-management.md), [Coding Agents](https://yomu.fyi/topic/coding-agents.md), [Golden Path](https://yomu.fyi/topic/golden-path.md), [Trust Calibration](https://yomu.fyi/topic/trust-calibration.md)

## Related reading

### [How Salesforce Scaled Engineering AI Adoption Without Stifling Innovation](https://yomu.fyi/post/how-salesforce-scaled-engineering-ai-adoption-without-stifling-innovat.md)
- Company: [Salesforce](https://yomu.fyi/company/salesforce.md)
- Author: Paul Constantinides
- Published: Nov 10, 2026
- Type: Problem & solution

Salesforce’s Hyperforce Runtime Platform (HRP) organization faced scattered AI experiments across 500 engineers in 70-80 scrum teams, with overlapping efforts and tight delivery commitments. The team established a federated AI adoption model with an AI Steering Committee, AI Champions, and the HRPDX shared Claude Code plugin to distribute AI instructions and skills. Teams contributed capabilities by solving their own pain points, with ownership assigned to the building team for end-to-end delivery and support. Within three months, AI adoption approached 100%, yielding a 26% gain in team throughput, a 32% improvement in inner-loop efficiency, and cycle time cut from 22 days to 11 days. The AI paved path model sustains adoption by continuously connecting emerging bottlenecks to available capabilities through metrics and a feedback loop.


### [Rider 2026.3: .NET 11 Support and JetBrains Air](https://yomu.fyi/post/rider-2026-3-net-11-support-and-jetbrains-air.md)
- Company: [Jetbrains](https://yomu.fyi/company/jetbrains.md)
- Author: Michael Hawker
- Published: Nov 10, 2026
- Type: Announcement

Rider 2026.3 ships alongside .NET 11 with complete support on day one, including C#15 features such as unions, closed hierarchies, and extension indexers. The release introduces JetBrains Air, an agentic development experience for directing and orchestrating agents and verifying their work inside JetBrains IDEs. Rider now supports new agent skills, and agents can execute tests through the Rider MCP. Game development improvements include new Lua and Verse plugins, professional debugging for Godot, and C++ debugger enhancements like natvis and command-line argument presets. A new Welcome screen allows users to get started without opening a solution, and the IDE can copy local project configuration into new Git worktrees for parallel workflows.


### [The State of Developer Ecosystem 2026: A Decade of Understanding Developers](https://yomu.fyi/post/the-state-of-developer-ecosystem-2026-a-decade-of-understanding-develo.md)
- Company: [Jetbrains](https://yomu.fyi/company/jetbrains.md)
- Author: Nadia Lokot
- Published: Nov 10, 2026
- Type: Announcement

The 2026 State of Developer Ecosystem report, based on 28,945 developers across 198 countries, marks its tenth edition. AI agents are now standard, with over half of employed developers using them in more than 60% of work tasks. Senior developers are harder to impress, as 60% of junior developers find AI solutions better than their own versus 47% of veterans. Python has overtaken JavaScript as the most widely used language at 65%, while TypeScript adoption has more than tripled since 2017. PostgreSQL gained nine percentage points to reach 59%, and 75% of developers would choose the same career again despite a tough job market.


### [MCP or CLI? Yes.](https://yomu.fyi/post/mcp-or-cli-yes.md)
- Company: [Sentry](https://yomu.fyi/company/sentry.md)
- Author: David Cramer
- Published: Nov 10, 2026
- Type: Announcement

Sentry ships a unified experience that bridges CLI and MCP, arguing both can coexist rather than one replacing the other. The CLI handles filesystem tasks like grepping raw event logs from session replays, which MCP cannot do efficiently without server-side pre-aggregation. MCP provides native tool discovery and steering that the CLI lacks, allowing agents to discover behavior without explicit instructions. The implementation shares authentication by launching the MCP server in stdio mode with a shim that passes CLI credentials through a different token header to bypass OAuth constraints. This approach works around current MCP limitations, including inconsistent support for deferred tool calling, instructions metadata, and events across different harnesses.


### [The State of Go in 2026: What Changed In the Golang Ecosystem Since Last Year](https://yomu.fyi/post/the-state-of-go-in-2026-what-changed-in-the-golang-ecosystem-since-las.md)
- Company: [Jetbrains](https://yomu.fyi/company/jetbrains.md)
- Author: Dominika Stankiewicz
- Published: Nov 10, 2026
- Type: Benchmark

The 2026 JetBrains survey reveals Go developers are experienced, with 66% having six or more years of coding experience versus 49% overall, and they earn higher salaries. The standard library net/http now leads web development at 46% usage, while TypeScript became Go’s top companion, rising to 43% in 2026 from 24% in 2024. AI coding agents are heavily adopted, with 86% of Gophers using them in the CLI at least weekly, yet 79% always review AI-generated code before accepting it. Testing is widespread, with 82% of Go developers writing tests for most or all projects, and go fix helps keep codebases modern. Major language updates in 2026 include generic methods, encoding/json/v2 with stricter defaults, UUID support in the standard library, and the Green Tea garbage collector.


### [Using AI to Build AI: How Agentforce Engineers Scale Automation Without Sacrificing Reliability](https://yomu.fyi/post/using-ai-to-build-ai-how-agentforce-engineers-scale-automation-without.md)
- Company: [Salesforce](https://yomu.fyi/company/salesforce.md)
- Author: Fereydoon Vafaei
- Published: Nov 9, 2026
- Type: Problem & solution

The Agentforce team at Salesforce faced the challenge of using AI to help build and maintain the Atlas Reasoning Engine without compromising reliability. The central risk was that AI could diagnose problems, modify the engine, test fixes, and review its own work while being confidently wrong. To address this, the team built an integrated engineering knowledge base using semantic retrieval to connect fragmented knowledge across incidents, fixes, and discussions. They implemented layered AI evaluation with an agent-as-a-judge model and human oversight, using different models to review work from different perspectives to avoid self-preference bias. As a result, typical bug-resolution time dropped from two to four weeks to one or two days, and system availability reached 100% in weekly reports during the observed period.
