【Monetize every AI moment】
Websites monetized pages.
Search monetized queries.
Social monetized attention.
AI is going to monetize intent.
But here's the problem — AI agents still don't have a native way to do commerce. Today's ads, affiliate links, feeds, redirects, and manual settlement flows? They were built for the old web. Not for AI conversations.
That's what AON — Agent Offer Network — is building.
We connect high-intent user conversations to relevant merchant offers, then handle everything behind the scenes:
→ intent matching
→ ranked offers
→ native rendering data
→ click tracking
→ conversion attribution
→ automated commission settlement
Developers keep full control of their product experience. Merchants reach users right at the moment of decision. Users get recommendations that are actually useful, contextual, and actionable.
One API call between user intent and revenue.
Not ads injected into AI.
Not another affiliate network.
The offer layer for the agent economy.
Monetize every AI moment.
#AIAgents #AgentEconomy #AICommerce #AIMonetization #AgenticAI
98 → 110 👀
Tiny goal, but I’m going for it.
No growth hacks.
No weird follow-for-follow stuff.
Just BUILDING AON, sharing what we learn, and meeting more builders here.
12 MORE TO GO 🚀
Another week of pushing aon_pro forward.
A big focus this week was reliability: tightening the protocol, release checks, and the overall path toward the next version.
We also made progress on offer retrieval and ranking, while continuing to improve the developer experience across the platform.
More of the work now is less about adding new pieces, and more about making the whole system stable enough to ship.
That part is taking more work than I expected.
#BuildInPublic #AIAgents
Last week, a few pieces started to land. This week, aon_pro feels a little more real.
v0.3 is getting closer, the recommendation flow is improving, and more of the tracking + offer infrastructure is falling into place.
We’re also pushing forward on MCP + Skills so more developers can actually find and try AON.
One thing I keep thinking about while building this: getting the infrastructure to work is only half of it. Making offers feel useful inside an agent, instead of interrupting the experience, is probably the harder part.
Curious how other agent builders are thinking about this.
#BuildInPublic #AIAgents
Some of last week’s “in progress” work is starting to land.
This week:
• the offer callback upgrade is wrapped up
• offer_full_data now has its own table
• AON-side e-commerce integration testing is done
• webhooks are built and waiting to go live
• the stats API is also getting ready to ship
We’ve also finished the first round of travel offer research and integration prep, while multi-goal / bid alignment and multi-language static offers are still in progress.
On the GTM side, we’re starting AON’s MCP registry + Skills submissions this week.
Feels good to see more of the stack moving from “working on it” to actually getting ready for use.
#BuildInPublic #AIAgents
A few things moved forward on AON this week.
You can now:
• handle multiple offers in a single callback
• use the upgraded PG / FS offer callback flow
We also added a separate offer embedding table, while full offer data support and multi-language static offers are still in progress.
E-commerce integration is moving along too — the aon_pro side is around 80% through integration testing.
More of the offer flow is starting to work end to end.
#BuildInPublic #AIAgents
One thing becoming clearer while building aon_pro:
Making an agent take action is getting easier.
Figuring out what happened after that is the harder part.
What was the user actually trying to do? Which offer mattered? And what should count as the result?
Anyone else running into this?