What's a good company AI adoption rate?
hms.harvard.edu is at 91.9%. This is 48.2pp above the community median (43.7%)..
91.9%
↑48.2pp above43.7% Community Median
Keep codifying prompts and monitoring adoption so the lead over peers is sustainable.
MARCH 2026
hms.harvard.edu Engineering AI Productivity Report
A focused summary of AI adoption, productivity lift, and code quality for the hms.harvard.edu engineering team.
See how AI-active teams rank this week on the Exceeds Leaderboards.
The hms.harvard.edu engineering team reports 91.9% AI adoption, 1.10× productivity lift, and 27.1% code quality across recent work.
These metrics track how AI integrates into delivery pipelines, how throughput changes when assistance is used, and the health of AI-supported code review outcomes.
What this report measures
We analyze commits and diffs to estimate AI adoption, productivity lift, and code quality for your engineering organization.
How to interpret these metrics
Use these signals to understand how AI assistance fits into day-to-day development, where enablement efforts drive throughput, and how review practices keep quality steady.
AI Adoption Rate
HIGH
91.9%
AI assistance is present in 91.9% of recent commits for hms.harvard.edu.
AI Productivity Lift
MODERATE
1.10×
AI-enabled workflows deliver an estimated 10% lift in throughput.
AI Code Quality
LOW
27.1%
Review insights show 27.1% overall code health on AI-supported changes.
How is the hms.harvard.edu team performing with AI?
The hms.harvard.edu engineering team reports 91.9% AI adoption, translating into 1.10× productivity lift while sustaining 27.1% code quality. These outcomes suggest AI-supported reviews are embedded in day-to-day delivery without trading off reliability.
Manager Questions Answered
Real questions engineering leaders ask about AI productivity, with live benchmarks and company-specific data.
Does AI actually make developers faster?
hms.harvard.edu operates at 1.10×. This is 0.03× below the community median (1.13×)..
1.10×
Roughly in line1.13× Community Median
Instrument reviewer assignment and AI summaries to trim the slowest merge steps and edge past the median.
How does AI affect code quality?
hms.harvard.edu holds AI-assisted quality at 27.1%. This is 3.9pp above the community median (23.2%)..
27.1%
Roughly in line23.2% Community Median
Invest in AI-specific test checklists and shadow reviews to keep quality slightly ahead of peers.
How evenly is AI use distributed across our team?
AI impact is concentrated—82.1% of AI commits come from a few experts, raising enablement risk.
82.1%
Run prompt-sharing sessions, codify AI review checklists, and incentivize broad participation.
How can I prove AI ROI to executives?
hms.harvard.edu has a solid ROI signal with room to strengthen either adoption, lift, or quality before presenting to executives.
—
Document case studies where AI accelerates delivery while maintaining quality, and expand playbooks across teams.
See how your full organization compares
Unlock personalized insights across all your repositories, teams, and contributors.
Securely connect Exceeds with your codebase to get commit-level insights on AI adoption and performance.
How Your Company Ranks
See how top engineering organizations compare across AI adoption, productivity lift, and code quality.
AI Adoption
% of commits with AI assistance
?
Companies in this quartile:
ID
idesie.com
(2904.2%)
IN
inngest.com
(1429.6%)
PR
prefeitura.rio
(87.4%)
NA
naduni.local
(87.4%)
Top 25% of teams adopt AI in 65-75% of their commits.
Productivity Lift
Cycle-time improvement vs baseline
?
Companies in this quartile:
AC
acad.pucrs.br
(1.12×)
MC
mcornholio.ru
(1.12×)
FL
fluxys.com
(1.01×)
TE
testinprod.io
(1.01×)
Top performers sustain 1.5× cycle-time improvements over six months when embedding AI into workflows.
Code Quality
Post-merge defect rate
?
Companies in this quartile:
IN
inngest.com
(701.7%)
ID
idesie.com
(649.2%)
GZ
gzgz.dev
(20.0%)
GW
gwu.edu
(20.0%)
Top 25% maintain quality above 92% while expanding AI usage, pairing automation with rigorous guardrails.
Rankings based on aggregated Exceeds AI dataset of 1.2M commits across open-source and enterprise engineering teams (Q4 2025).
Top contributors
Top contributors combine high AI adoption and quality output. Encourage internal sharing of best practices.
Contributor
Commits
AI Usage
Productivity Lift
Code Quality
SL
SEHI L'YI
Commits6
AI Usage20.0%
Productivity Lift1.50x
Code Quality20.0%
CF
Cesar Ferreyra-Mansilla
Commits39
AI Usage94.0%
Productivity Lift1.15x
Code Quality76.0%
DM
David Michaels
Commits126
AI Usage92.0%
Productivity Lift1.13x
Code Quality20.0%
BM
Bianca Morris
Commits102
AI Usage92.0%
Productivity Lift1.11x
Code Quality20.0%
SA
sarahgonicholson
Commits146
AI Usage92.0%
Productivity Lift1.06x
Code Quality20.0%
Encourage knowledge transfer from top AI users to others through internal mentoring or recorded "AI coding walkthroughs." Balanced adoption across the team typically improves overall performance by 12-15%.
Cross-Organization Network
Shared Repositories
4Activity
184 Commits16
14
19
37
42
20
15
21
0
0
0
0
LessMore
Your Network
9 PeopleBianca-Morris
Memberaschroed
Membercrfmc
Memberdmichaels-harvard
Memberngehlenborg
Membersarahgonicholson
Membersehilyi
Membertkakar
MemberTDeSain
MemberWhy these metrics matter for engineering managers
Faster delivery
1.4x lift → predictable roadmaps
Safer velocity
93% quality → lower rollback risk
Equitable gains
AI less dependency on heroes
Governance
Depth monitoring audit-ready
