The Early Scale: Amazon Business Hits $60B in Annualized Sales, and Agentic AI Is Coming for the Procurement Stack

Good morning

The machines are spending money faster than the humans managing them. Amazon Business just crossed $60 billion in annualized sales, OpenAI is embedding engineers inside enterprise clients, and Wall Street is literally putting its post-trade infrastructure on a blockchain. The common thread: every major B2B platform is betting that AI agents will soon do the buying, the deploying, and the settling, and whoever owns that pipe owns the relationship. If you run or market a B2B business, the question is no longer "should we think about AI?" It's "do our products even show up when an agent goes shopping?"

The Big Three

Amazon Business Hits $60B in Annualized Sales, and Agentic AI Is Coming for the Procurement Stack

Amazon Business has reached $60 billion in annualized gross sales, cementing its place as the dominant B2B marketplace, according to Digital Commerce 360. But the bigger story is what comes next: agentic AI systems that autonomously discover, evaluate, and purchase on behalf of enterprise buyers. Deloitte research cited by Digital Commerce 360 warns that B2B suppliers who fail to optimize for machine-readable product data risk becoming invisible to the next generation of procurement tools.

The B2B angle: Audit your product catalog now for machine-readable attributes, structured pricing, and API accessibility before agentic buyers start bypassing suppliers with messy data.

OpenAI's $150M Enterprise Push Puts Engineers Inside Client Operations

OpenAI launched a standalone professional services firm and a $150 million partner program specifically designed to close the gap between AI model access and real enterprise deployment, the Wall Street Journal reports. The move embeds forward-deployed engineers directly inside client operations, a strategy lifted from Palantir's playbook. It signals that selling AI access alone is not enough: the battle for enterprise AI spend is now fought at the implementation layer.

The B2B angle: Enterprise technology vendors should expect OpenAI-affiliated teams competing for the same implementation budgets previously owned by traditional system integrators.

Tempus AI's $1.7B Personalis Deal Headlines $7.4B Half-Year Surge in Digital Health M&A

Tempus AI is acquiring cancer genomics firm Personalis for $1.7 billion, the largest single deal in a wave of health-tech consolidation that saw $7.4 billion in digital health venture funding in the first half of 2026 alone, according to Modern Healthcare citing Rock Health data. CMS also just created a dedicated health technology office, signaling federal alignment with AI adoption. For health system operators, the vendor landscape is consolidating fast.

The B2B angle: Health system procurement teams should reassess multi-year vendor contracts now, before smaller point-solution providers get absorbed or pivoted by acquirers with different roadmaps.

Also worth knowing

Okta acquired AI security startup Permiso for approximately $200 million, per TechCrunch. Permiso specializes in detecting identity threats across cloud environments, a capability that becomes critical as machine identities now outnumber human ones 144 to 1. For enterprise security buyers, expect Okta to fold these detection capabilities into its core identity platform within the next product cycle.

Oracle signed a 10-year software contract with the Pentagon worth up to $7 billion, CNBC reports. It is one of the largest enterprise software deals in federal history and signals that AI-capable cloud and data platforms are now core defense infrastructure, not optional IT upgrades.

Industrial manufacturing M&A hit a record $173 billion over the past year, up 28%, per a PwC report cited by Manufacturing Dive. Roland Berger separately forecasts 6 to 9 percent annual automation capital spending growth through 2030. Operators sitting on aging equipment should treat this consolidation wave as a forcing function, not background noise.

Smart plays for the week

Do a product data audit this week: check that every SKU in your B2B catalog has structured attributes, machine-readable pricing, and clean API output before agentic procurement tools make messy data a dealbreaker. Digital Commerce 360 reports that Amazon Business's agentic AI era is already arriving, and Deloitte warns that suppliers with unstructured data will become invisible to autonomous buyers.

If you sell to health systems, get in front of your procurement contacts now to reconfirm roadmap alignment before your vendor gets acquired and pivoted away from your use case. With $7.4 billion in digital health funding in just the first half of 2026 and Tempus AI's $1.7B Personalis deal, the health-tech vendor map is reshuffling faster than most multi-year contracts anticipated.

B2B marketers: map your demand-gen content to account intelligence signals by integrating your ad engagement data with a platform like Demandbase or 6sense so sales gets CRM-ready contacts, not just click reports. Reuters reports that Demandbase and DemandWorks just integrated to automate pipeline activation from account signals, and 6sense is launching AI-recommended leads in August, making manual lead hand-off a competitive disadvantage.

Something to think about

This framing from the Wall Street Journal's reporting on OpenAI's enterprise push captures the core B2B challenge right now: having the best model means nothing if the implementation layer is broken. Every enterprise vendor, not just AI companies, should be asking whether they're selling access or delivering outcomes.

Teach me something: Agentic AI Procurement

Traditional e-commerce assumes a human buyer who searches, compares, and clicks. Agentic AI procurement flips that: a software agent acts on behalf of the buyer, autonomously researching options, evaluating them against preset criteria, and placing orders, all without a human in the loop. For B2B suppliers, this matters enormously because agents don't browse the way humans do. They query structured data feeds, APIs, and catalogs. If your product data is messy, incomplete, or locked behind a non-API storefront, an agentic buyer will simply skip you and move to a competitor whose catalog is machine-readable. The Amazon Business $60B milestone is being reported alongside this shift precisely because Amazon's infrastructure is already designed for agent-to-agent commerce.