Ramp
Ramp is an American financial technology company specializing in spend management solutions for businesses. Its core products include corporate credit cards, expense reimbursement management, bill payment, and financial automation services, designed to help companies control costs and improve financial efficiency. Ramp primarily serves small and medium-sized enterprise customers, simplifying routine corporate financial workflows through the integration of software and financial services.
Timeline (last 90 days)
According to LogRocket's October tool rankings, Cursor's share of Ramp spend fell from 41% to 26%, while Claude Code and OpenCode rose.
Ramp opened up Glass, an AI workspace built around sharing AI agents and workflows across teams.
A user on Reddit shared a case claiming that Ramp's product team, with the help of Glass, went from about 300 manually logged feature requests per month to 20,000 automatically captured per month.
Ramp Glass represents the trend of AI agents moving from personal tools toward organizational infrastructure, and its significance is akin to the leap from personal Excel spreadsheets to shared Google Sheets.
Ramp's latest AI Index shows token spending declined 5.2% month-over-month, while token usage is rising.
Ramp's economists noted that the decline in token spending is almost entirely driven by competition between OpenAI and Anthropic, with open-source models still accounting for less than 5% of enterprise spending.
Ramp built a browser-based QA agent called Testo that launches 100 combinations based on real production data to simulate real user actions, catching 425 bugs in the past 30 days.
Ramp's code review tool ReviewBuddy now automatically handles 93% of PRs, allowing top engineers to focus on the remaining 7%.
Ramp's Inspect has accumulated millions of sessions, with 75% of PRs built by it, and 1,000 PRs submitted by non-engineers in the past month.
Ramp developed a coding agent called Inspect that runs in Slack, responds within 5 seconds, and can return interactive deployment previews.
40 more timeline events
All Facts (20)
Ramp团队采用分层审查的人机协作模式，AI作为代码审查的第一道防线，人类审查者聚焦架构合理性、业务逻辑正确性和安全合规等高层次判断
80%UnverifiedAccording to Ramp data, 80% of enterprise revenue for OpenAI and Anthropic comes from just 1% of customers.
60%UnverifiedRamp is an enterprise spend management platform that provides services including corporate credit cards, expense management, and procurement.
60%UnverifiedAccording to LogRocket's October tool rankings, Cursor's share of Ramp spend fell from 41% to 26%, while Claude Code and OpenCode rose.
50%UnverifiedRamp Glass represents the trend of AI agents moving from personal tools toward organizational infrastructure, and its significance is akin to the leap from personal Excel spreadsheets to shared Google Sheets.
50%UnverifiedA user on Reddit shared a case claiming that Ramp's product team, with the help of Glass, went from about 300 manually logged feature requests per month to 20,000 automatically captured per month.
50%UnverifiedRamp's latest AI Index shows token spending declined 5.2% month-over-month, while token usage is rising.
50%UnverifiedRamp's economists noted that the decline in token spending is almost entirely driven by competition between OpenAI and Anthropic, with open-source models still accounting for less than 5% of enterprise spending.
50%UnverifiedRamp built a customer insights agent that uses ETL and data pipelines to pull data sources from across the company, combined with vector retrieval and clustering to understand products, teams, and features.
50%UnverifiedA context window of one million tokens represents less than 0.5% of Ramp's entire Gong call records.
50%UnverifiedFrom June 2025 to May 2026, Cursor's share of enterprise AI coding tool spending tracked by Ramp fell from 41% to 26%, and JetBrains data showed workplace adoption rates declining from 18% to 12%.
50%UnverifiedThe conversation content comes from a Ramp × GPT-6 Astra demo video on YouTube and represents a vendor-perspective demonstration.
50%UnverifiedRamp uses computer use technology to fetch receipts from merchant sites in the background for customers, but requires extensive manual prompting to guide the model's actions on various merchant websites.
50%UnverifiedOne of Ramp's core value propositions is simplifying accounting workflows through automation, and computer use is key to achieving this.
50%UnverifiedElevenLabs initially integrated financial automation platforms such as Ramp to address bill payment issues, and subsequently expanded to other financial modules.
50%UnverifiedDuring Ramp's data hydration process, the same email address may represent multiple companies, so fuzzy matching is required
50%UnverifiedRamp built a persistent execution mechanism based on Temporal, representing each tool call and model call as an activity that forms a durable thread, which can resume execution from the point of interruption if a worker unexpectedly exits
50%UnverifiedRamp converts unstructured sales data (call records, emails, notes) into vectors via an embedding model and stores them in a vector database to support semantic retrieval
50%UnverifiedRamp uses a Postgres transactional layer to maintain transactional guarantees and referential integrity between entities, and retains complete metadata
50%UnverifiedRamp uses DBT together with the Snowflake cloud data warehouse for offline batch processing computation, and then flows data back through reverse ETL
50%