---
id: "lib-2026-11-08-moving-from-ai-adoption-to-an-ai-operating-model"
type: "library"
title: "Moving from AI adoption to an AI operating model: How Docusign engineering went beyond tool rollout"
lang: "EN"
org: "The New Stack (Docusign engineering)"
source_type: "Article"
date: "2026-11-08"
week: "2026-W45"
url: "https://nosov.ai/library/lib-2026-11-08-moving-from-ai-adoption-to-an-ai-operating-model/"
original_url: "https://thenewstack.io/docusign-ai-engineering-adoption/"
tier: "C"
tier_name: "Weak signal"
tier_reason: "A Docusign-sponsored New Stack piece that relays Docusign's own engineering-blog figures with CTO quotes; adoption and code-volume numbers only, no delivery or quality outcomes."
summary_basis: "full-text"
original_minutes: "9"
tags: ["enablement", "business transformation", "champions model", "adoption"]
---

# Moving from AI adoption to an AI operating model: How Docusign engineering went beyond tool rollout

Tier C (Weak signal): A Docusign-sponsored New Stack piece that relays Docusign's own engineering-blog figures with CTO quotes; adoption and code-volume numbers only, no delivery or quality outcomes.

Summary based on the full text.

## In short

A New Stack post (8 November 2026) by editor-in-chief Nick Lucchesi, marked "Docusign sponsored this post", on how Docusign moved its 1,100+ engineers from tool rollout to shared agents. Usage stalled at 49% until a six-person champions team and then about 75 embedded "Multipliers" took it to 93%. CTO Sagnik Nandy says adoption is only the start and the real measures are feature velocity, reliability and customer satisfaction, but no such figures are given.

## Key numbers

- 49%: engineers actively using AI coding tools in June 2025, when a reported 18% productivity gain had levelled off
- 65%: active use by October 2025 after a six-person Eng AI Champions team
- 93%: active use within two months of embedding about 75 AI Multipliers in teams
- approximately 52%: rise in lines of code updated per developer from the starting point
- 10 of 12: configuration files an agent modified correctly on first pass in about 15 minutes, vs about a week by hand

## What it reports

- June 2025: 49% active use; a reported 18% productivity improvement had levelled off.
- Six-person Eng AI Champions team built guardrails and ran workshops; adoption reached 65% by October 2025.
- About 75 AI Multipliers in development teams deployed shared solutions and fed problems back; active use reached 93% within two months.
- Lines of code updated per developer rose about 52%; the article notes this does not establish faster delivery or better quality.
- Internal tools: Kody (coding), Chronicle (observability), DocuWiki (code-aware docs), and Krew, a platform where teams publish reusable agents ("Krewmates") that others can copy.
- An overnight agent searches error logs and drafts fix PRs for engineers to review in the morning.
- Guardrails: no code merges unreviewed; agents cannot merge a PR or take irreversible actions without approval.

## How they got it

Interview with the CTO plus figures taken from Docusign's own engineering blog. No measurement method is given.

## Limits

- Sponsored by the company it covers; figures are self-reported and second-hand.
- Adoption and code volume are activity measures, not outcomes.
- The 10-of-12 file example is a single task and still needed correction.

## Why it matters for companies

The structure is the useful part: a small central team sets practice, embedded people in each team carry it and report back, and successful agents are published so others inherit them instead of rebuilding. Treat the adoption curve as a distribution result. The open question, which the article itself raises, is whether delivery speed and quality moved, and that needs measuring before the next round of investment.

Original: https://thenewstack.io/docusign-ai-engineering-adoption/