AI adoption · Navigara

Who on your team uses AI,
and how deeply?

Navigara reads your AI tools’ own usage data and scores every engineer on five habits. One level each. One next step each. No survey.

Start free trial
curl -fsSL https://navigara.com/cli | sh

Run it on a laptop. First adoption data in seconds.

dashboard · ai adoption
AI adoptionHow people spread across the five levels, from all five habits over the trailing 90 days at each point. People without AI activity in that window are left out.
Illustrative team · 28 people · 12 weeks
Week 43, by level
171055
  • Curious 1
  • Regular 7
  • Fluent 10
  • Delegator 5
  • Orchestrator 5
Team average: Regular → Fluent over 12 weeks.
Team average= 1 person in 1 week

Trusted by engineering teams

Kiwi.comFinshapePartners BankaPurple TechnologyGreysonItrinityESETFTMOKeggSecond Foundation
01Five habits

Adoption is five habits, not a feeling.

Claude Code, Codex, Cursor and Copilot report what they do: every prompt, every agent minute, every tool call. Navigara turns that into five habits per person, each with a target. Hands-off is this engineer’s lowest habit, so it is the next step.

Five habits, each against a targetTap a habit to open it · illustrative
  • %

    How the score is earned. A picture, not this engineer’s data.

    What it is

    Agent minutes per prompt before a person steps in. Short prompts with constant corrections are pairing; a long run is delegation.

    Agent minutes per prompt1.2 min of 6 min target

    Do this first

    Give the agent a whole task in one prompt. Now 1.2 min, target 6.

    Then try

    • Write the acceptance criteria into the prompt, so the agent knows when it is done.
    • Let the agent run the tests and fix its own failures before it comes back to you.
One engineer, trailing 90 days. A habit at its target scores 100%; below it, in proportion. Read from the AI tools’ own telemetry, recomputed every day.Score = the average of the five
02One level

Five scores average into one level.

Each level is a fifth of the score, over the trailing 90 days, recomputed every day. A team’s level is the average of its active people: one person is one vote. Coach, do not rank.

  1. 1Curious0–20%Tries AI now and then.
  2. 2Regular20–40%Uses AI most weeks, mostly for quick questions.
  3. 3Fluent40–60%AI is a daily habit.
  4. 4Delegator60–80%Hands whole tasks to agents and lets them finish.
  5. 5Orchestrator80–100%Runs several agents at once, and work goes on while away.
03One next step

Every engineer gets a next step, not a rank.

The next step is the habit with the lowest score. This Fluent engineer uses AI 5.2 days a week, but their prompts run 1.2 minutes before they step in (target: 6). Their own page shows that one habit and what to do first, not a ranking of the team.

my performance · ai adoption
How far along is your AI adoption?
3Fluent59% · trailing 90 days
FluentNext: Level 4 · Delegator

One engineer, one page. It shows their own level and one habit to work on, never a leaderboard.

Your next step · Hands-off
1.2 min of 6 min

Give the agent a whole task in one prompt. Now 1.2 min, target 6.

Then try

  • Write the acceptance criteria into the prompt, so the agent knows when it is done.
  • Let the agent run the tests and fix its own failures before it comes back to you.
The next read-out shows whether it moved.
04Does it pay?

Put the level next to the work it delivered.

Levels show habits, not output. So the adoption page plots every team by level and by delivered work per person. In this illustrative company Payments, the highest-level team, delivers 8.1 ETV per person a week; Data delivers 3.4. Then the price of that work, per team and per week, is one click away in AI spend, and the capacity it added in AI ROI.

ETV is the unit of delivered work on the performance page. Illustrative teams, not customer results.

Which teams lead on AI adoption and output?Average level and delivered work (ETV) per active person a week, by team, over 90 days. Dot size: people.
Illustrative company · 4 teams
  • Payments8.1 ETV / person / weekDelegator · 105 people · $84 per ETV
  • Platform6.2 ETV / person / weekFluent · 82 people · $71 per ETV
  • Mobile5.0 ETV / person / weekFluent · 93 people · $96 per ETV
  • Data3.4 ETV / person / weekRegular · 70 people · $128 per ETV
ETV: Navigara’s unit of delivered engineering work, scored on every merged change. Levels show habits, not output; this is where you check whether the two move together.
05Start

One command. First adoption data in seconds.

curl -fsSL https://navigara.com/cli | sh
Start free trial
01

One command per laptop

Installs the Navigara CLI, signs in, uploads past Claude Code and Codex sessions and streams new ones. Cursor and Copilot report live, with no history upload. A whole team: the same installer through your MDM.

02

Usage within minutes, habits the next day

AI spend shows up in minutes. The five habits and the first dots appear the next day; a level needs 90 days of history, which the Claude Code and Codex upload brings with it.

03

No source code leaves the machine

The tools send usage metadata: sessions, models, cost, tokens, durations, tool names. Prompt text is optional and can be switched off. Read-only everywhere.

Developer AI adoption questions.

What does developer AI adoption measure?

Five habits per person, read from the AI tools’ own telemetry: how often they use AI, how long agents run, how much they hand over per prompt, how many agents run at once, and how long agents run with nobody steering. Each habit is scored against a target, over the trailing 90 days, and recomputed every day.

Does higher developer AI adoption mean better productivity?

Only if delivered work moves too. The adoption page plots every team by level and by ETV per active person, and the AI spend page prices that work per team and per week. When usage climbs and delivery stays flat, you see it on the same screen.

→Start here

See your team’s
AI adoption by tomorrow.

Free trial, one command on each laptop, no survey. Where Claude Code or Codex history is already on the laptop, every person gets a first level the next day.

Start free trial