Campfire

Campfire is an AI-native ERP for growing finance teams, covering general ledger, multi-entity consolidation, revenue recognition, reporting and close management on one platform. Founder and CEO John Glasgow started it in San Francisco in 2023 (High). It is venture-owned. A $65M Series B co-led by Accel and Ribbit was announced on October 15, 2025, about 12 weeks after a $35M Series A.

Crunchbase puts total funding at about $103.5M. Revenue is not disclosed. The Series B post claims 10x revenue year to date (Medium). The LinkedIn company page shows 51 to 200 employees (High, logged-in sweep 2026-10-14). Bottom line: a well-funded AI-native ERP climbing from mid-market close and GL into enterprise migrations, still closed-source and quote-only.

What does it sell?

The homepage headline is “The AI-Native ERP” with “One platform for general ledger, revenue, reporting, and close management.” Six marketed parts: core accounting (GL, journals, chart of accounts, dimensions, budgets, allocations, multi-entity consolidation), revenue (contracts, subscriptions, usage revenue, ASC 606), close (checklists, flux, reconciliations, accruals), Ember as the assistant and agent layer, Accounting Intelligence built on a proprietary model called LAM, and a platform claiming 100+ integrations and 1,200+ permissions (Medium on the integration count).

The target is mid-market and enterprise finance teams replacing legacy ERPs, with NetSuite and SAP as the named incumbents. Geography is US-first, with a global compliance claim that is not independently verified (Medium). Named customers are PostHog, Decagon and Replit. Full billing often stays with a partner such as Tabs. Pricing is not published. The pricing page returns a 404 (High).

What does it ship for AI agents?

Shipped: Ember, the in-product assistant and agent layer, and an MCP server with 12 documented tools covering chart of accounts, transactions, entities, vendors, departments, tags, the three statements, budgets, contracts and aging. All 12 are reads and reports. No MCP write or proposal tools are documented (High). An October 7 LinkedIn post notes a listing in the OpenAI ChatGPT app directory. LAM carries a vendor claim of 95%+ accuracy on key accounting tasks with no public evaluation set (Medium). The chat layer wraps commercial models alongside LAM (High on the pattern).

What does a buyer need to check?

• Reviews: G2 about 4.8 on roughly 28 reviews. Capterra not checked.
• Top complaint: features still developing and rapid product change (Medium).
• Close-time claims: “5-6 Days Shaved off close” and “20x Revenue Without increased headcount” are vendor-sourced with no third-party methodology (Medium).
• Pricing: contact sales only, with no published entry price.
• Ledger controls: this review reports API endpoints that delete or reopen journal entries, which conflicts with an immutable-books pitch (Medium until the live OpenAPI is re-read).
• Compliance: SOC 1 and SOC 2 claimed in secondary coverage, to confirm on the live trust page (Medium).
• Ownership stage: venture-backed, Series B, hiring across departments as of October 2026.

What to watch

• A published pricing page, at campfire.ai/pricing.
• MCP write or proposal tools, at docs.campfire.ai.
• The live OpenAPI confirming or removing delete journal-entry endpoints, in the API reference.
• A named enterprise migration off SAP, on the blog and in case studies.
• A break in the Tabs partnership or a deeper native billing launch, on integrations pages and the changelog.
• An independent LAM evaluation or model card, in the trust center or docs.