The state of engineering, 2026

AI made your team faster.
The code wait got worse.

Copilots and agents write code faster than teams review, test and release it. The research is in: more output, longer review queues, more instability. Agile Analytics measures where that shows up in your loop — from the tools you already run.

Book a demo →See the platform
+98%
PRs merged, high vs. low AI adoption · Faros 2025
+91%
longer review time, same teams · Faros 2025
30%
little or no trust in AI code · DORA 2025
Tickets in flight Stuck > 24h AI-assisted output Shipped
build flow · illustration
AI assist
Plan
12
Code
28
bottleneck
Review
41
QA
19
Ship
8
The Faros 2025 pattern: more output, longer queue at review.illustration · not customer data
02 · What the research says

AI doesn't fix a team.
It amplifieswhat's already there.

DORA’s 2025 conclusion, from close to 5,000 respondents. Faros AI measured the same thing in telemetry. Same tools, same engineers — the difference is whether anyone can see where the extra output goes.

● Faros AI · 2025 · 10,000+ developers

More output. Longer queues.

Teams with high AI adoption, compared with low. At company level, the correlation with better results was gone. A year later, on 22,000 developers, Faros found incidents per PR had tripled and 31% more PRs merged without any human review.

PRs merged
+98%
Tasks completed
+21%
Review time
+91%
PR size
+154%
Bugs per developer
+9%
● What Agile Analytics measures

Measure where it lands.

Output is measured where it is typed. Value is measured where it ships. Agile Analytics reads Git, Jira, CI/CD and monitoring and puts the two side by side, per team:

  • FlowLead time, deployment frequency, Sprint Insights
  • Quality gatesChange failure rate, MTTR, Error Budgets
  • MaintenanceFeature vs. non-feature work, per sprint
  • PeopleDevEx surveys, interruptions, Kudos

Also in the picture: GitClear (211M lines, 2020–2024) saw duplicated code rise from 8.3% to 12.3% of changed lines and refactoring fall from 25% to under 10%. METR’s randomised trial found experienced developers 19% slower with early-2025 tools while believing they were 20% faster — a result its authors warn not to generalise, cited here for the gap between feeling and measuring. GitClear 2025 · METR 2025

03 · How we see it

One loop. Three lenses.
One question: did it ship?

Engineering productivity isn’t a single number — it’s a flow. AgileEx, DevEx and OpsEx are the three lenses that make the whole loop measurable, so AI output is counted where it lands rather than where it was generated.

All lensesAgileExDevExOpsEx
illustration · one sprint
PlanCodeBuildTestReleaseReviewRefineBacklogDeployRunObserveAlertResolvePostmortemSLOLearnENGINEERINGone loopTHREE LENSESAgileEx · DevEx · OpsExbottleneck
04 · Quality gates for AI-assisted code

Speed is easy to see.
Quality has to be measured.

Four questions every team building with AI should be able to answer from data. Three of them Agile Analytics answers today; the fourth is on the roadmap, and we say so rather than pretend.

Does output pile up at review?measured today

Lead time for changes and deployment frequency per team, from Git and CI/CD. Sprint Insights shows how much of each sprint actually ships as features. Review wait and PR size as first-class metrics are next.

Does stability hold?measured today

Change failure rate and MTTR from your deploys and incidents. Error Budgets and SLOs from your monitoring, with alerts when a budget burns. This is where DORA’s “higher instability” shows up first.

Is maintenance eating the sprint?measured today

Sprint Insights classifies Jira tickets as feature or non-feature work, so the share of each sprint that goes to maintenance rather than features is visible per team — and whether it moves as AI output grows.

What does the AI spend buy?roadmap

Seats plus metered usage run $200–600 per engineer per month across tools (DX, 2026), and almost nobody sets that against what shipped. Cost per team next to lead time is on our roadmap. Until it is measured, it is a workshop conversation, not a dashboard.

05 · How it works

Connect. Baseline. Compare.

No new process, no self-reporting. The data already exists in the tools your teams use every day.

ConnectStep 1

Access tokens you create and can revoke, for GitHub, GitLab or Bitbucket, Jira, your CI and your monitoring (CloudWatch, Google Cloud Monitoring). Slack for Kudos, if you want it.

BaselineStep 2

The first sprint gives every team its starting point: lead time, deployment frequency, change failure rate, MTTR, the feature-to-maintenance ratio.

CompareStep 3

Teams that lean on AI against teams that don’t. This sprint against last. The gates against their targets. The question “did it ship?” gets a per-team answer.

06 · Example · a team’s gates

Engineering productivity
has an SLA now.

A team sets targets for lead time, change failure rate and review turnaround, and Agile Analytics tracks every breach. The values below are an example of what that looks like — not customer data.

Lead time p50on target
4.2d
target ≤ 5d · example
Change failure rateat risk
12%
target ≤ 10% · example
Review turnaround < 24hbreached
62%
target ≥ 95% · example

Stop guessing.
Start measuring.

Connect Jira, your Git host and your CI. Your first sprint report shows where AI output waits, whether the gates hold, and how much of the sprint shipped — from your own data.

Sources