> ## Content Index > Fetch the complete content index at: https://www.implicator.ai/llms.txt > Use this file to discover other available public pages before exploring further. # Ramp Opens Glass Early Access After a $1.5 Million Week of AI Spending - URL: https://www.implicator.ai/ramp-glass-ai-workspace-token-spending/ - Published: 2026-10-15T18:06:41.000Z - Updated: 2026-10-15T18:06:41.000Z - Description: Ramp is opening early access to Glass after a $1.5 million week of AI spending. Its shared workflows and project meters aim to make company knowledge reusable, while the launch figures leave outside buyers without an independent measure of results. How much of that experience will carry over? - Author: Marcus Schuler - Tags: AI News Ramp opened [early access to Glass](https://ramp.com/blog/introducing-glass?ref=implicator.ai), its internal AI workspace, to outside businesses on October 13\. Co-founder and co-CEO Eric Glyman said Ramp had spent $1.5 million on AI tokens in one week in May. The workspace lets employees save useful workflows and company-specific definitions for colleagues to reuse. Ramp's pitch draws on its own use of Glass: shared work can reach colleagues instead of staying in individual chat histories. Glass also gives each project a spending meter and limits, with requests passing through [Ramp Router](https://ramp.com/blog/router-launch?ref=implicator.ai) for model selection. Tokens are the chunks of text AI models process, and their usage contributes to the bill. What Changed - Ramp opened Glass early access on October 13, offering outside businesses a workspace for reusable company workflows. - By the launch, 323 Ramp employees had contributed more than 6,500 edits to its shared knowledge. - Glyman reported a 30% bill reduction from Router, a separate claim from independently measured Glass savings. - Ramp's launch announcement supplied internal usage examples without a broad external-customer productivity audit. AI-generated summary, reviewed by an editor. [More on our AI guidelines](https://www.implicator.ai/about/). ## Shared workflows A small team built Glass to give colleagues the setups Ramp's most advanced AI users had assembled for themselves. Glass connects company knowledge across Linear, Notion, Google, Slack and Gong, and supports open and closed models. Colleagues can join the same session, allowing a support agent to bring an engineer into a ticket with the existing context. Glyman's launch example concerned counting U.K. customers. Finance had previously supplied a definition excluding U.S. companies with U.K. offices, so his query used that rule without another explanation. By the October announcement, 323 Ramp employees had contributed more than 6,500 edits to Glass's knowledge. The reusable part of that query was the definition Finance had already supplied. FREE WEEKDAY MORNING BRIEFING Follow the tools changing how teams use AI. The Implicator Morning Briefing filters the AI news cycle to the stories worth your attention and explains their consequences. From San Francisco, every weekday at 4:45 a.m. Pacific, 7:45 a.m. Eastern. Email address Send me tomorrow’s briefing Check your inbox for the confirmation link. About five minutes. No hype. No spam. In May, an account manager saved a prompt that extracted feature requests from customer calls. Monthly captured requests rose from about 300 manually entered in Salesforce to 20,000 by the launch. A prompt notified customers when requested features shipped, sending more than 5,000 emails by the announcement. Those are Ramp's internal counts of captured requests and notifications. They do not establish request quality, hours saved or how many features were delivered. ## Project costs Router, publicly launched on August 19, selects models and service tiers and shows how requests were handled, including token usage and cost. Glass puts spending visibility alongside the work, so teams can set project limits before costs exceed them. The meter shows the cost of work while it is happening, giving employees a view of project spending before the invoice arrives. Glyman said Ramp's bill fell 30% overnight when it moved its work to Router. Glyman's launch post juxtaposed the $1.5 million May week with the savings but supplied no baseline dollar amount or comparison holding workloads constant. The figure describes Ramp's Router experience; it is not an independently measured savings result for outside Glass customers. ## Reusable skills and a support study Anthropic introduced [Agent Skills](https://www.anthropic.com/research/skills?ref=implicator.ai) in October 2025, packaging instructions, scripts and resources that could be shared and used across Claude products. Know someone who'd find this useful? [✉️ Email it to a friend in one click](mailto:?subject=A%20newsletter%20I%20think%20you%27d%20like&body=This%20is%20one%20of%20maybe%20three%20newsletters%20I%20actually%20read.%20The%20rest%20just%20pile%20up%2C%20unread%2C%20judging%20me.%0A%0AAnd%20yes%2C%20this%20email%20mostly%20wrote%20itself%2C%20which%20is%20a%20little%20on%20the%20nose%20for%20an%20AI%20newsletter.%20Doesn%27t%20make%20it%20wrong.%20implicator.ai%20is%20good.%0A%0ASubscribe%20free%3A%20https%3A%2F%2Fwww.implicator.ai%2Fsubscribe%2F%3Futm%5Fsource%3Dnewsletter%26utm%5Fmedium%3Dforward%26utm%5Fcampaign%3Demail%5Fforward), or they can [subscribe free here](https://www.implicator.ai/subscribe/?utm%5Fsource=newsletter&utm%5Fmedium=forward&utm%5Fcampaign=forward%5Fto%5Fcolleague). A [study summarized in June 2023](https://www.nber.org/digest/20236/measuring-productivity-impact-generative-ai?ref=implicator.ai) examined an AI assistant introduced mostly between November 2020 and February 2021 to roughly 5,000 support agents at a Fortune 500 software company. Resolved issues per hour increased nearly 14%. Less experienced and lower-skilled workers gained most, while top performers experienced zero or small negative effects. Analysis of conversations suggested the assistant helped spread experienced agents' practices to colleagues. That study measured a different assistant in customer support, not Glass. The Glass launch announcement provided no broad external-customer productivity audit. ## Financing and access On October 15, Ramp announced a [$1.85 billion primary financing](https://www.prnewswire.com/news-releases/ramp-reaches-60-billion-valuation-as-annualized-revenue-tops-2-billion-302909315.html?ref=implicator.ai) led by Dragoneer at a $60 billion valuation. Company figures as of October 1 put annualized revenue above $2 billion, with positive cash flow. Glass remains in early access. What results will businesses outside Ramp obtain from the shared workspace? Frequently Asked Questions What is Ramp Glass? Glass is Ramp's AI workspace for saving company-specific definitions and workflows that colleagues can reuse. It connects internal tools and lets coworkers join the same session. Who can use Glass? Ramp opened early access to outside businesses on October 13, 2026\. It is personally onboarding the first businesses. Why did Ramp build Glass? Eric Glyman described a $1.5 million week of token spending in May. A small team built Glass to give colleagues the setups Ramp's most advanced AI users had assembled. What does the 30% savings figure measure? Glyman said Ramp's bill fell 30% when it moved its work to Router. The launch supplied no baseline dollar amount or workload-controlled comparison, and the figure is not independently measured savings for outside Glass customers. Do the feature-request figures prove productivity gains? Ramp said captured requests rose from about 300 manual Salesforce entries a month to 20,000\. Those counts do not establish request quality, hours saved or features delivered. Its launch announcement provided no broad external-customer productivity audit. AI-generated summary, reviewed by an editor. [More on our AI guidelines](https://www.implicator.ai/about/). ## Related stories [Best GitHub Tools for Building Claude Code SkillsA Claude Code skill is a folder with a SKILL.md file. The GitHub tools around it now do distinct jobs, from authoring with skill-creator to versioned plugin distribution and multi-CLI sync. Here is which repository to reach for, and when, plus the security check every third-party skill needs.Implicator.ai![](https://www.implicator.ai/content/images/2026/06/2026-06-18-22.38.43-skills_toolbox@2x.webp)](https://www.implicator.ai/the-best-github-tools-for-building-claude-code-skills/) [OpenAI Says AI Saves Workers an Hour. Researchers Disagree.OpenAI claims ChatGPT saves workers an hour daily. MIT found most AI deployments return zero value. Harvard researchers coined "workslop" for AI output that looks professional but advances nothing. Whose data should enterprises trust?Implicator.ai![](https://www.implicator.ai/content/images/2025/12/2025-12-08-06.17.36-OpenAI_Office_Time_Workslop@2x.webp)](https://www.implicator.ai/openais-productivity-numbers-look-great-independent-researchers-arent-so-sure/)