Skip to main content
Live

575K+ AI events recorded on our own engineering fleet in the last 30 days.

Tracelet / Leadership

Understand your engineering AI rollout.

Review tool adoption, policy findings and machine coverage in one conversation with engineering and security. Then see what the spend bought, and how much of it rests on a single vendor.

TRACELET.AILeadershipIllustrative example
Leadership / overview

The engineering AI rollout, in one view.

Last 30 days
AI health index
82Healthy
  • Tool adoption Strong
  • Policy findings Reviewed
  • Machine coverage On track
  • Decisions reviewed Current
Composite score across the fleet
Value this month
$186Knet value, illustrative

1,420 hours saved against a $41K token spend this period.

Spend, hours saved and net value, fleet-wide
Policy findings
Policy decisions Recorded Open findings Review Machine coverage Tracked
Coverage by team
Teams Projects People
Enrolled machines by team, project and person.
Needs attention
Unregistered MCP servers to reviewTeams with machines not yet enrolledBlocked actions awaiting review
Items for the next engineering and security conversation.
AdoptionFindingsCoverageOne leadership view

Illustrative example: adoption, findings, coverage and fleet-wide value are available signals today.

Three questions for the next rollout conversation.

Every answer comes from enrolled machines rather than a survey: adoption from recorded AI activity, findings from policy decisions, coverage from what is actually reporting. Values are illustrative.

Benchmarks are agreed with each team, so the tick on every row is your bar, not ours. The rollout stage underneath places the organisation on the adoption guide, so the conversation starts from where you actually are.

Tracelet / ScorecardLast 30 days
01

Which teams are using AI coding tools?

Tool adoption
62%
Enrolled machines with recorded AI activity · illustrative
02

What did policy find?

Policy findings
91
Findings evaluated, severity-weighted, 28 days · illustrative
03

How much of the fleet is covered?

Machine coverage
72%
Machines enrolled and reporting · illustrative
Rollout stageFull maturity breakdown further down
ExploringAdoptingScaling52/100OptimisingTransforming

The fourth question: who does the rollout depend on?

Model and vendor strategy reads your own AI-provider spend and shows how much of it sits with one vendor, what a price change there would cost each month, and how much already has a same-tier alternative. Developer AI-tool spend gets its own cost-by-model table, kept apart from the provider bill so the two never blur.

Concentration is not a failure; it is a negotiating position. The leverage numbers exist so the next vendor conversation starts with figures rather than a feeling.

Tracelet / Model & vendor strategyLast 30 days
Diversification
38of 100
Single-vendor
Top vendor
71% of spend
10% price move there
$2.9K a month
Concentration and price-shock exposure, own provider spend
Spend by provider
  • Frontier models Primary vendor$29.1K71%
  • Frontier models Second vendor$6.6K16%
  • Open-weight models Self-hosted$3.3K8%
  • Embeddings and small models Two vendors$2.0K5%
$41.0K this period · Developer tool spend by model sits in its own table
Leverage

The three numbers to take into the next vendor conversation.

Blended rate
$6.40per M tokens

Weighted across every provider and tier in use.

Tiering opportunity
$7.2Ka month

If routine workloads moved to the standard tier.

Portable spend
58%of spend

Has an active same-tier alternative today.

ConcentrationExposureLeverageOne vendor view

Illustrative example: provider spend, concentration and leverage over the same 30-day period as the overview above.

How far along is the rollout, really?

Benchmarks and maturity place the organisation on the same five stages as the scorecard above, with the composite score behind the stage and the four pillars that feed it. The gap to the next stage is named, so the conversation is about the one thing to lift rather than the whole curve.

Reference bands are configurable placeholders agreed with your teams, not measured peer data. Values are illustrative.

TRACELET.AIBenchmarks & maturityIllustrative exampleLast 30 days
AI maturity stage
52of 100 · Scaling

8 points to Optimising. Security posture is the drag: 42 against a 70 bar.

Composite of the four pillars below · stage bands every 20 points
You vs target
  • Tool adoption Target 75%62%13 pts under
  • Code keep rate Target 70%58%12 pts under
  • Security score Target 704228 pts under
  • Spend per active developer Band up to $250$220Inside band
Reference bands are configurable placeholders, not measured peer data
Health pillars

The four scores behind the composite, each against its own bar.

Adoption
62of 100 · target 75

4 points up on the prior 30 days

Code quality
58of 100 · target 70

2 points up on the prior 30 days

Engagement
46of 100 · target 60

3 points down on the prior 30 days

Security
42of 100 · target 70

5 points up on the prior 30 days

StageTargetsPillarsOne maturity view

Illustrative example: composite maturity, core metrics against their targets and the health pillars behind them.

What is the fleet exposed to?

Risk and compliance rolls policy findings, violations, secrets and destinations into one fleet score, with the enforcement mode beside it so a low number reads as exposure recorded rather than damage done. Break-glass activity is the same keycards engineering approves on the governance side, rolled up for the board.

Every count comes from policy decisions on enrolled machines, over the same 30 days as the rest of the page. Values are illustrative.

Tracelet / Risk & complianceLast 30 days
Fleet score
42of 100
At risk
Enforcement
allow_all · recorded, not blocked
Weighting
Severity and age of open findings
Exposure across enrolled machines, 30 days
Open critical
4

of 34 critical + high findings

High severity
23

29 medium alongside

Secrets detected
6

across the 30-day window

Unapproved destinations
9

of 138 distinct domains contacted

Fleet score trend
+5over the last 30 days · 28 below the 70 bar
Weekly fleet score, higher is better
Findings by severity
  • Critical11
  • High23
  • Medium29
  • Low28
Open
31
Resolved
60
Total
91
The same 91 findings as the scorecard above
Top violated rules
RuleSeverityTotalOpen
Unapproved destinationMedium218
Sensitive path accessHigh146
Blocked tool callHigh93
Secrets exposureCritical62
Dangerous commandCritical52
Ranked by total violations · open counts feed the fleet score
Break-glass activity
Requests
18
Granted
72%
Pending
2

Time-boxed elevated access. The same keycards engineering approves on the governance side, rolled up for the board.

Median window 30 minutes · every grant names its approver
ExposureFindingsBreak-glassOne risk view

Illustrative example: fleet score, findings, violated rules and break-glass activity in one board-grade view.

Where are spend and return heading?

Strategic forecast projects the cycle from today's run-rate: spend, hours saved and net return, with headcount and unit-price levers to test the plan before the budget round or the next vendor conversation.

Run-rate projections are for planning, not commitment. Values are illustrative.

TRACELET.AIStrategic forecastIllustrative exampleDay 18 of 30
Projected monthly spend
$44.2K

$1.47K a day run-rate

Spend to date
$26.5K

12 days left in the cycle

Projected net ROI
$198K

1,540 hours saved, projected

Active developers
186 / 300

62% adoption

Spend forecast
Actual, days 1–18 Projected, days 19–30 Today
Cumulative spend this cycle · linear run-rate from today
Scenario planner
Headcount
−20%Flat+20%+50%
Unit price
−10%Flat+25%
Scenario monthly spend
$44.2K Matches baseline
Scenario net ROI
$198K Matches baseline

Assumes a linear run-rate for the rest of the cycle at today's blended rate. A planning scenario, not a commitment.

Move a lever and both figures move with it
Run-rateScenariosReturnOne planning view

Illustrative example: run-rate, projected spend and return, and a scenario planner over the current billing cycle.

Design partners

Start the conversation. One call to establish fit and how we would work together.

Become a design partner