Three announcements, one story

In June, three AWS announcements landed within three weeks. Most people read them separately. They’re one story.

June 10: Swami Sivasubramanian published how “frontier teams” across Amazon redesigned software development around AI — median 4.5x productivity, measured across hundreds of teams.

June 12: Francessca Vasquez showed AWS Professional Services applying that method to itself — a multi-agent delivery system on live engagements, timelines compressed from months to days.

June 30: a $1 billion commitment to Forward Deployed Engineering — thousands of engineers embedding the proven method inside customer teams, with a partner program to scale it further.

Prove it on yourself. Turn it into a method. Commit at scale. In that order.

After a month of writing about FDE, that sequencing is what I keep coming back to. Anyone can announce a number. The discipline of earning it internally first — hundreds of teams, then your own consulting arm, then customers — is what makes the number credible.

My read on where enterprise AI goes from here: the winners won’t be measured by pilots launched or models benchmarked. They’ll be measured by systems running in production, and by how much more capable their customers became along the way.

That’s the era AWS just committed a billion dollars to. I like our odds.