Solutions Factory

Three products that share one runtime and one bill.

APG (Agents Proving Ground) runs the agents. CRM and Drengur are the two departments built on it. Bought together, they share one runtime, one inference meter and one invoice. Our delivery service builds agents around your operation.

One meter

One runtime, one meter, one invoice.

Bought together, the products share one runtime and one inference meter, and arrive on one invoice. Spend stops at a cap you set.

One runtime
The engine runs the agents behind the CRM and Drengur.
One meter
Model use is counted once, across the suite, and itemized.
One invoice
A flat software fee, and inference as its own line.
Spend for one runDemo data

Spent so far

$0.50

Stopped at the cap
$0.00Cap $0.50

The agent stopped at $0.50. Nothing more is spent until a person raises the cap.

Invented numbers. A real run shows its real cost on the bill.

The software is yours to run. The agents do the work.

The industry has started calling this service-as-software. The agent does work a person used to do, and you pay for the work. There is no charge for the right to log in. It is a useful name because it explains the price model without an argument.

Per-seat pricing counts humans. When agents are the ones doing the work, counting humans measures the wrong thing. So the suite is priced as a software fee plus inference metered as a utility, and the meter is printed on the bill.

Drengur, one inquiry at a time.

A client email goes in. A checked itinerary comes out and waits for you. Pricing comes from dated supplier rate sheets, and nothing is sent without a person approving it.

Demo data. An invented client email moves through eight stages: client email in, intake, itinerary plan, pricing, logistics check, voice pass, quality gates and itinerary PDF. Pricing uses dated supplier rate sheets and leaves a line open rather than inventing a price. The finished PDF waits for the travel designer to approve it and is never sent automatically.

Coming next, or by engagement.

The case room and the brain are coming. The personal assistant is built by engagement. None of them is part of the suite today.

See what is coming →Talk to us about an engagement →

Bought together, the products are worth more together than apart.

One runtime

Every agent runs on APG. When something misbehaves there is one place to look, with one set of cost caps and one audit trail.

One bill

A software fee for what you license, plus inference itemized once across all of it. You are not juggling three subscriptions and three opaque credit pools.

One security posture

Three-layer tenant isolation and the same confirm discipline everywhere, so a review covers the whole estate.

One exit

The runtime is yours. The software keeps running on your hardware for as long as your license stands, and letting a retainer lapse does not stop it.

Where this sits against the alternatives.

Solutions Factory suite compared with hosted AI-helper platforms and self-hosted agent tooling.
FeatureSolutions FactoryHosted helper platformsSelf-hosted agent tooling
What you getDepartments built for your operationThe same prebuilt helpers as everyone elseA canvas and the work to do on it
Where it runsYour infrastructureTheir platformYour infrastructure
Inference costItemized per token; metering through us, with the markup stated once it opensBundled into a tierYours to buy and manage
Who is accountableWe operate it under retainerA support ticketYou are
Time to first valueWeeks. It is a build.Same dayDepends on your team

Hosted helper platforms win on speed, and we will not pretend otherwise. The trade is generic and immediate against specific and yours.

Start with one. Add the rest when it earns it.

Most engagements start with one product or one agent and grow because the first one worked. A discovery call is where we work out which one that should be.