the rails are live — UCP · AP2 · llms.txt
Your next customer is software. We make sure it can find you, read you, and choose you — and we keep it that way as the rails move.
01/the shift
Agents research, compare and shortlist on their owner's behalf. Increasingly, the first visitor to your product page isn't a person — it's software acting for one, and it forms its opinion before a human ever sees you.
02/the consequence
ChatGPT, Claude, Perplexity and AI Overviews don't return ten links. They return one answer. If they can't read you, you're not in it — and there's no page two to fall back to.
03/the work
Structured, answerable, transactable. We make your catalogue, your claims and your policies readable by machines — and we keep them that way as the rails keep moving.
03 — what we do
The first three run today. The fourth is being built now, in the open — because the rails it needs are still setting.
01live
Nobody buys what they can't find.
Discovery across both surfaces that matter now — classic search and generative answers. The structure, schema and content that earn a citation rather than a link.
02live
Being found is not being picked.
Continuous optimisation for both readers. Humans get a page that persuades. Agents get claims, specs and policies they can parse without guessing — because a guess is a lost sale.
03live
The market moves while you sleep.
Benchmarking and monitoring that never stops — competitor pricing, positioning shifts, and what the answer engines are telling people about you this week.
04in development
Agents won't just read. They'll buy.
Agent-to-agent commerce across the emerging rails — UCP, AP2 and whatever follows. We're building this now so you're ready before it becomes table stakes.
04 — the array
Named, scoped and built to hand work to each other. This is how that work moves between them.
opportunity & trend hunter
Watches the agentic-commerce frontier and writes the weekly signal.
market & pricing mapper
Benchmarks catalogues against the field and models true market pricing.
monitoring
Brand mentions, price moves, compliance drift, and what the engines answer.
content engine
Turns findings into pages built to be cited, not just ranked.
agentic readiness
Scores whether machines can find, read, trust and transact with you.
agent-facing commerce
Answers and sells on your behalf, agent to agent, across the emerging rails.
five in production · envoy in development · the rails it needs are still setting
05 — proof
Scale of work rather than revenue — those are the numbers we can stand behind, and they’re the ones that say what we can actually do.
Modelled true market pricing across the category, then drove the supplier renegotiation that followed.
→COGS down 40–80%
Co-developing a platform on AWS — intake, physician routing, EMR, e-prescribing — and running the HIPAA programme across all of it.
→AI policy-drafting agent + live tracking
Brand identity, SEO and GEO content, social presence and paid acquisition for a venture selling across two markets.
→Built from zero
A positioning engine scoring SEO, GEO and agentic readiness — productised as the scanner further down this page.
→Run it on your own site
06 — the auditor
Fifteen checks across three layers, scored 0–100 each. The scan is free and the findings come back written by a person — because a score on its own tells you nothing about which problems cost money.
The classic layer. Still the majority of discovery, and still where most of the fixable damage is.
Generative engine optimisation. ChatGPT, Claude, Perplexity and AI Overviews return one answer, not ten links — this decides whether you're in it.
The newest layer and the least covered. Reading you is not the same as being able to buy from you.
the finding nobody expects
If your site blocks our request, that isn’t an error in the scan. It’s the result.
It’s exactly what every shopping agent, answer engine and research bot sees when they come looking — and they don’t email you about it. They just move on to a competitor who let them in.
Free. Written findings within 24 hours.
08 — about
Your autonomic nervous system runs your heartbeat, your breathing, your digestion. You never decide to do any of it. It runs underneath, continuously, and it only reaches your attention when something is wrong.
That’s the standard we hold commercial operations to. Research that doesn’t wait to be asked. Monitoring that doesn’t sleep. Content that keeps pace with a market moving nightly. The work currently eating your week should be running underneath it.
Autonomic Commerce is led by Rex Denger. The past year went into doing the work rather than describing it — benchmarking a regulated catalogue against forty competitors and rebuilding its pricing from the floor up; running a HIPAA programme across a hundred and eighty federal controls with an AI agent drafting the policy; standing up brand, search and paid acquisition for a cross-border property venture that started with nothing.
The through-line was the same every time. The parts that looked like they needed a team turned out to need a system.
So that’s what gets built. We are not claiming commerce already runs itself — nobody can. The rails only finished setting this year. What exists today is the Auditor, Scout’s weekly brief, and a set of systems already doing real work on real catalogues. The rest is being built in the open, week by week, and you’re welcome to watch it happen.
09 — request an audit
Tell us where to look. You’ll get a written answer — not a sales call, not a calendar invite, not a deck.
Rather just email? hello@autonomiccommerce.com