Primary
Turns meeting notes and email into assigned tasks under the right client.
What runs today
Throughline runs on ten production systems plus an agent platform, built by a team of two or three, running every day inside real operating companies.
None of these was built to sell. Each one started as a problem inside a company we run. The weekly numbers arrived late, so we built Scorecard. Client history lived in one person's head, so we built Pensieve. Anomaly Detector exists because ad spend needed a daily check and nobody had the day.
The agent platform is the shared plumbing the ten run on, built the same way, for the same reason. Most started as a search for something to buy. The off-the-shelf tools didn't know our meeting rhythm or our client list, so building won. That's our whole build-versus-buy answer for AI: buy when software already fits the work; build when the work is specific to your business.
The proof
One system of ten.
The rebate-and-margin tool we built inside Kelsan has recovered over $100,000 in margin the manual process was missing. Kelsan, a multi-state distributor in our own group, runs it every morning inside its Epicor P21 ERP. The tool uses AI to read the rebate terms in the contracts and check every line item against them, whether anyone has time that day or not.
Operations
Five systems run the operating side of our companies: tasks, weekly numbers, meetings, client memory, ad spend. Scorecard is still a plain table nobody's had time to make pretty.
Turns meeting notes and email into assigned tasks under the right client.
Pulls the week's numbers together before the meeting starts.
Preps the weekly leadership meeting and carries every open item forward.
Keeps each client call and note, and briefs whoever needs the history.
Watches ad spend daily and flags the numbers that moved.
The marketing side
The last four do marketing work. One of the companies these systems grew out of is a marketing agency, so that's where they came from. They're here for the pattern. The same shape of build that publishes an article on schedule can run the margin report your sharpest person only gets to when there's time.
Every system here started as one chore worth fixing. The audit is where yours gets found and priced.
Where they run
One fact to weigh: every one of these ten runs every day inside a real operating company, not on a demo account. We don't publish a company's name without their written say-so, so where naming one would help, we hold it until they clear us.
The one reach outside our own walls so far is Autoblogger, which does its work for clients outside our own companies. We'll name them when they say yes.
The next step
Every system here started as one chore worth fixing. The AI Capability Audit is where yours gets priced: fee, terms, and what you keep, on one page.
Fixed fee ($7,500)fixed durationa build-or-don't-build verdict you keep either way