AI adoption

Token bills tell you what you spent.Origin tells you what you bought.

Origin reconstructs agent activity from the endpoint, so you can see what agents were asked to do, the steps they took, and what they produced.

Request Demo →
Usage

See what’s becoming part of daily work.

Before you renew a license or roll a tool out further, check whether people kept using it. Week-by-week usage of each agent, MCP server and skill shows which ones people tried and which they now use regularly.

Origin / Explore / Active sessions

Interactive example
Aug 10 → Oct 4
  • Claude Code1,940+373%
  • Codex860+617%
  • Cursor710+11%
  • Microsoft 365 Copilot300−21%
  • Gemini CLI140New
  • Ollama44+10%
Work

Understand what agents are working on.

Origin sorts every session by the work inside it, from code review to contract redlines.

Origin / Workspace / Kinds of work

Interactive example

Weekly sessions, change since Aug 10

Aug 10 → Oct 4
Aug 10 level+84%
Aug 10Aug 31Sep 28

Agents

  • Claude Code44%
  • Claude37%
  • Codex19%

Tools

Snowflakexlsxsql-style-guide

Spend

Follow the work behind your AI bill.

Finance will ask what the bill paid for, and a token count can’t answer it. Break estimated spend down by kind of work, then open the sessions behind it to show what agents were doing.

Origin / Explore / Token volume

Interactive example

≈ $33,200 est.3.7B tokens

Sep 7 → Oct 4
Sep 7Sep 14Sep 21Sep 28
  • Fix the race in the billing webhookClaude Code≈ $6.80
  • Review the auth refactor pull requestCodex≈ $4.10
  • Add retry logic to the invoice exporterClaude Code≈ $3.40
Output

Show what the spend produced.

Open a session to see what it produced, from changed files to merged pull requests, beside its estimated cost and every step the agent took to get there.

Explore the platform

Origin / Sessions / X-ray

Interactive example

Add retry logic to the invoice exporter

Claude Code148 turns

Cost
≈ $3.40 est.
Pull request merged
1
Files
4
Linear issues
4
  • Pull request #2214Retry failed invoice exports with backoffMerged
  • 4 pushesfix/invoice-retry
  • CI2 runsPassed
  • Issue #2198Invoice export drops rows on timeoutClosed
Reuse

Turn repeated work into a skill.

One person works out how to get an agent through a task, and the next person starts from scratch. Origin spots the workflows that keep repeating and drafts each one as a skill for your team to review and refine.

Origin / X-ray / Workflows

Interactive example

Collect a week of merged pull requests, group them by area, and draft customer-facing notes.

Candidate skill (draft)

name: release-notes-from-prs
description: Use when drafting release notes from a range of merged pull requests.
  1. List pull requests merged since the last tag
  2. Group them by product area
  3. Write one plain line per change
  4. Flag any line that names an internal ticket

uses: GitHub, Linear

Members

  • Draft release notes for 4.12
  • Summarize what shipped since 4.11
  • Release notes for this week’s merges
Research

What usage reports leave out.

AI Adoption and the Distribution of Intelligence Problem

Vendor dashboards show AI token spend by user but not whether the work warrants it. The prompts and agent trajectories that answer that live on the endpoint.

Vendor dashboard

Token consumption by userShown

What the tokens were spent onNot shown

The invoice tells you how much, not why.

Read the post

A PDF Picked My Model

We told Claude Code to use Fable and handed it a one-page biology PDF. The summary came back from Opus. You paid for Fable and got billed for a blend of Fable and Opus.

Claude Code · told to use Fable

Read bio-handout.pdf and write a one-paragraph summary

Answered byOpus

Sent to Fable but answered by Opus.

Read the post

Know what your AI budget is buying.
Back what’s working.

Every AI program gets asked what it spent, where the money went and what it produced. We’ll walk you through how Origin answers all three.

Request Demo →