Ijeoma Onuosa

Builder, FinTech since 2016. Nonbank treasury. EECS x ORFE x Economics @UMich Mortgages, Portfolio Mgmt, RegTech, World history 🚜🛢️🌾🌽 #JPY #VXX #QQQ 🇺🇸

[Ijeoma Onuosa @CapitalBanker]

In the beginning... there was work 🙏🏾

[Ijeoma Onuosa @CapitalBanker]

@katedeyneka @imanionuosa

[Ijeoma Onuosa @CapitalBanker]

@DilumSanjaya @imanionuosa

[Ijeoma Onuosa @CapitalBanker]

@dashboardlim FREE

[Ijeoma Onuosa @CapitalBanker]

The cubicle-based jobs—customer service, data entry, payroll processing—created a vital ladder to the middle class, helping replace factory work lost to overseas competition.

Now, these white-collar jobs are fading, too, thanks to continued offshoring and, increasingly, AI ...

[Ijeoma Onuosa @CapitalBanker]

@elonmusk dynamic pricing enters the chat

[Ijeoma Onuosa @CapitalBanker]

@adcock_brett 👀

[Ijeoma Onuosa @CapitalBanker]

@paulg 🎯

[Linas Beliūnas @linasbeliunas]

Wild: Anthropic has not only passed OpenAI in reported revenue run rate. At ~$45 billion ARR, Anthropic may now be the fastest-growing software business of all time 😳

To put that in perspective, Salesforce did roughly ~$38B in FY2025.

Anthropic is reportedly annualizing more than that while most companies are still debating whether AI agents are production-ready.

The numbers are absurd:

→ ~$10M ARR in late 2022
→ ~$1B ARR by Jan 2025
→ ~$14B ARR by early 2026
→ ~$45B ARR by May 2026

That is not a normal SaaS curve. That is a software company learning how to sell software that writes software.

And Claude Code is the center of the story.

↳ In under a year, it reportedly hit a ~$2.5B run rate.
↳ Enterprise customers spending $1M+/year allegedly doubled from ~500 to ~1,000 in just 2 months.
↳ And more than 80% of Anthropic’s revenue is now enterprise.

While others chased consumer virality, Anthropic went straight for enterprise codebases, recurring budgets, developer workflows, and work teams can’t just “turn off.”

That is the monetization gap.

Fewer users than competitors, but dramatically higher spend per customer.

Sure, Anthropic will spend an obscene amount on training, inference, and compute.
But if revenue keeps scaling like this, the old SaaS math starts to break.

We thought AI would be a feature inside software.

Instead, it’s replacing parts of it.

If this trajectory holds, the biggest winners in AI won’t be the apps.

It will be those who own the A layer everything else runs on.

And right now, that layer is ClaudeOS.

[Ijeoma Onuosa @CapitalBanker]

@linasbeliunas 🎯

[Ijeoma Onuosa @CapitalBanker]

@natolambert agreed

[Ijeoma Onuosa @CapitalBanker]

the most powerful AI models are doubling in capability every four months

[Ijeoma Onuosa @CapitalBanker]

@ifeosaa 🏆

[Ijeoma Onuosa @CapitalBanker]

@SoulzBTC TRADING

[Ijeoma Onuosa @CapitalBanker]

@deedydas 🏆

[Ijeoma Onuosa @CapitalBanker]

@alifcoder Handbook

[Ijeoma Onuosa @CapitalBanker]

@BoucherNicolas AI

[Ijeoma Onuosa @CapitalBanker]

@alifcoder Ebook

[Ijeoma Onuosa @CapitalBanker]

@BoringBiz_ nice observation

[Ijeoma Onuosa @CapitalBanker]

@AfricanPatriot_ @marcportermagee 🎯

[Ijeoma Onuosa @CapitalBanker]

@Samaytwt 💯