Scheduling tools spent thirty years getting better at drawing the bar chart.
We spent ours making the logic visible.
A project is a combination of processes: best practices, methods, quality procedures — sometimes just the idea in someone’s head. Those processes live at every level of the schedule, macro to micro, and each earns its place by what it gives to the process above it. And in its pure form — the way you would draw it in BPMN — every process loops: do, check, redo until it’s right. A pure process assumes no resource limits and no deadline. Left to itself, it loops forever.
The schedule is what makes it stop. A schedule network cannot hold a loop — the logic must run forward — so every loop has to be unrolled into a number of passes, and choosing that number is the act of agreement. A report updated three times. A design reviewed twice. An inspect-and-repair cycle with an agreed repeat count, priced into the budget. Those numbers are not guesses; a mature quality process knows how many passes its work needs. Agreed loops force competent work — the opposite of “fake it until you make it” — and they give change a fair basis: new scope adds a loop, and compensation is assessed against the loops everyone signed.
The logic is what gives the dates their meaning. The work of project planning is the work of translating one into the other — visibly, defensibly, so every stakeholder can stand behind every date. And the schedule, like the process it serves, keeps improving — traceable, flexible, and compliant through every revision.
Six principles, applied every day.
Six beliefs from the field. They shape every schedule we touch — and every tool we build.
Logic first. Dates follow.
Most scheduling tools show links; almost none show logic. A link says A comes before B. Logic says why, who decided it, and what it would cost to change — the justified record the professional bodies have asked for all along. Without it, the only defence left is gut feeling: good enough in your head, worthless in a dispute room. We work at the logic level, because when the logic is right, the dates defend themselves.
The fundamentals scale down.
Scope, dependencies, realistic durations: the industry writes these into megaproject frameworks and lets everything smaller run on instinct. We apply the same discipline at every size. A two-week outage deserves real logic as much as a multi-year programme.
Fit for purpose, not maximum.
A schedule can fail by being too heavy as well as too weak. Detail beyond the decisions it serves turns a project team into schedule administrators. We size the schedule to the decisions it must support — no more, and no less.
Days, not scores.
A risk-matrix score is deniable; a consequence in days, computed from your own schedule logic, is not. We put risk in days and paths, then go straight to actions, owners, and need-by dates. Undeniable numbers end the avoidance.
AI for the work we don’t want. Never the work we don’t know.
AI drafts, types, and tidies — it takes the drudgery, not the judgment. A person authors the schedule and a person signs it, and every AI suggestion must show its evidence before it earns a place.
The record outlives the project.
Good planning leaves a record, and the record is an asset. The same logic graph that runs today’s schedule teaches your juniors, folds your quality processes into the plan, and captures the good practice your best people carry in their heads. It becomes the foundation your future projects — and your own AI, under your control — can build on. From the site manager to the project director to the next bid: one integrated record, more valuable every time it is used.
Three stages of how you’ll use it.
“I finally see it.”
First time you load a real schedule. Hidden dependencies surface. False critical paths reveal themselves. The shock of seeing what was always there.
“We must record this.”
Once your team has seen it, they cannot un-see it. The next change comes with a K-Flow note. The next review has an audit trail. Decisions stop disappearing.
“We cannot defend without it.”
By the third project, the audit trail is the asset. The forensic record. The bid evidence. The reason the next contract was won — or defended.
One record. Two readers.
The industry has discovered that AI without a structured, typed record can only guess — and guessing, in a schedule, is expensive. So a question is now being asked everywhere: how do we keep our process models understandable to people while making them useful to AI? We think that is one property, not two. What makes a record clear to a human — this link exists for this reason, decided by this person, at this cost to change — is exactly what makes it computable by a machine.
Others teach AI to guess what happened. We record why it happened — so AI never has to guess. The typed, justified record your team builds in its daily work becomes the ground your own AI, under your control, can stand on.
Test the worldview against your data.
The fastest argument for what we believe is what LogicReader does to your schedule. Try it.