See which teams use Claude Code, what it costs per team, and what it produces: sessions, tokens, accepted edits, commits and pull requests. Worklytics turns Anthropic's APIs into team-level adoption, output and cost metrics, without reading any code.
.png)
.png)
.png)
The problem
Anthropic's own analytics show Claude Code usage and spend. They can't tell you which teams get value, how Claude Code changes delivery, or how it compares with Cursor, Codex and GitHub Copilot.
Usage and cost are not tied to your HR structure, so you cannot see which departments, managers or locations justify their seats.
A quick fix and a long agent run look the same in a usage total. They need different responses.
If you also run Cursor, Codex or GitHub Copilot, there is no single view of your AI coding tools.
A vendor dashboard cannot show how much of the working day involves AI, or how it changes focus time, meetings and output.
See who adopts Claude Code and why, what it changes in the engineering week, and what it costs by team.
What Worklytics adds
Anthropic reports on its own product. Worklytics connects Claude Code to your org structure, your other AI coding tools and the engineering work itself.
Active users, active days and sessions for Claude Code by team, manager, location and tenure, from your HR data.
Compare each team with the rest of your organization and with peer companies (benchmarks in beta).
Fluency goes beyond logins: how often and how intensively people use AI, and AI penetration, the share of available working time that involves AI.
See which teams have made AI part of daily work and which are still experimenting, and track movement after training.
Lines added and removed, commits and pull requests attributed to Claude Code, and accepted versus rejected edits, by team.
Shows whether Claude Code output is being kept and shipped, not just generated.
Find what predicts AI use in your organization: whether a person's manager uses AI, tenure, role and team norms.
Aim enablement where it will move adoption most, starting with managers.
Compare heavy and light Claude Code users, before and after rollout, on pull request throughput, review cycle time, focus time and meeting load.
Shows whether AI is changing the work itself, not just adding another tool.
Tokens, models and cost for Claude Code by team and per active user, so you see where spend goes before the bill arrives.
See where spend goes and what it returns before you renew or consolidate tools.
Worklytics vs. native analytics
Anthropic's Claude Code analytics cover usage and spend well. These are the gaps for teams that run more than one AI coding tool or need to connect AI use to the work.
| Capability | Claude Code analytics | Worklytics |
|---|---|---|
| Seat activation & weekly active users | ✓ | ✓ |
| Sessions and active days per user, trended | Partial | ✓ |
| Token cost by team and per active user | ✗ | ✓ |
| Code output by team | ✗ | ✓ |
| Cross-tool view (Cursor, Codex, GitHub Copilot) | Claude Code only | All AI tools unified |
| Breakdowns by manager, tenure and location | Not tied to HR data | ✓ |
| Impact on focus time, meetings, collaboration and output | ✗ | ✓ |
| AI penetration of the working day | Not available | By team, trended |
| User-level pseudonymization (no raw data) | Basic aggregation | SHA-256, minimum group sizes |
How it works
Worklytics reads Claude Code usage through the Claude Admin API and Claude Enterprise Analytics. No agent, no SDK and no code changes, and nothing is written back to your organization.
Personal identifiers are hashed with SHA-256 before analysis, and results are shown only for groups above a minimum size. Prompt and response text is never stored. With the Worklytics proxy, this happens before data leaves your environment.
Connect your HRIS, calendar or GitHub to see Claude Code use by team and next to pull requests, reviews, meetings and focus time. Claude Code works on its own too.
Adoption, fluency, task, impact and cost views come pre-built. Most teams see metrics within a week of setup, with trends building over the following 30 days.
Privacy & Compliance
Worklytics measures AI use without reading employee content. It connects through admin-level APIs only, is audited annually against the SOC 2 standard, and is privacy-first by design.
The data is already there. Worklytics connects through read-only APIs and gives IT, HR and leadership pre-built views, each with a clear action to take.
