Why the Term Matters Less Than the Function
I spent 10 years managing construction projects before I built Provan. During that time, I watched project teams run multi-million dollar contracts out of spreadsheets, Outlook inboxes, and the personal memory of whoever happened to be the lead PM that month. Not because they were lazy or behind the times. Because no tool on the market actually solved the problems they faced every day.
Document management systems store documents. Field management platforms track inspections and site activity. Project scheduling tools produce Gantt charts. None of them tell you that a Construction Contracts Act payment schedule response is due in three working days, that a Special Condition amended the standard notice period, or that the contractor's latest variation claim conflicts with a previous Engineer's instruction.
An intelligent operating system does not manage your project for you. It makes sure you see everything you need to see, when you need to see it, so your professionals can make better decisions with complete information.
Monday Morning: What Hits Your Dashboard
You arrive at your desk. The system has already processed the weekend's correspondence, any overnight emails to the project inbox, and updated all deadline calculations. Your daily priority view shows:
- Two CCA payment schedule responses due this week, with the response window closing dates and the claimed amounts
- A notice of potential claim received from the contractor last Thursday. Under the amended Clause 13.3.1, you have 10 working days to respond. The system shows the deadline and links to the original Special Condition that created the amendment
- A variation instruction issued two weeks ago that has not yet been priced by the contractor. The standard Clause 9.8.3 gives 10 working days. That window closes Wednesday
- One subcontract where payment claims have exceeded assessed value by more than 15% for three consecutive months, flagged as a financial risk indicator
None of this information is new. It all existed in the project files. The difference is that without the system, someone had to remember to check each of these items, track each deadline manually, and connect the dots between correspondence received last week and obligations created months ago in the contract conditions.
Six Domains, One View
The system operates across six integrated domains. Each one covers a different aspect of project and business performance. Here is what each domain actually does during a typical project week.
Pipeline. For PMC firms managing their business development alongside project delivery, the Pipeline domain tracks live tenders, upcoming opportunities, and follow-up actions. On a Monday morning, it might show that a tender response is due in 12 days and the capability statement needs updating with the current project's latest milestone. For more on building those capability statements, see our article on winning construction tenders.
Contracts. This is the core. Every obligation from every contract has been ingested, including all Special Conditions amendments. The system knows that this particular NZS 3910 contract has 147 Special Conditions. It knows which ones modified standard timeframes. It knows which created new obligations that do not exist in the standard form. It tracks every deadline and links every piece of correspondence to the relevant obligation.
Projects. Programme tracking, milestone monitoring, and progress reporting. The system compares actual progress against the baseline programme and flags deviations. When the electrical subcontractor falls two weeks behind, it calculates the downstream impact on related trades and alerts the project team before the delay cascades.
People. Team workload, capability mapping, and key person risk monitoring. If your lead contract administrator is managing obligations across three projects simultaneously, the system flags the concentration risk. It also ensures that if that person is unavailable, another qualified team member can access the full obligation register and pick up the work. For more on this, read our article on why AI does not make decisions and how the system enhances rather than replaces your professionals.
Finance. Payment claim tracking, variation value exposure, and financial forecasting. The system shows the total value of instructed but unpriced variations, the gap between claimed and assessed payment values, and cash flow projections based on the current payment schedule.
Risk. Cross-cutting risk intelligence that pulls signals from every other domain. A subcontractor showing declining progress, inflated claims, and delayed correspondence responses gets flagged as a financial risk before anyone has to ask the question. A head contract obligation approaching its deadline without a recorded response gets escalated.
The value is not in any single domain. The value is in the integration. A subcontractor payment pattern in Finance connects to a programme deviation in Projects connects to a risk flag in Risk. No spreadsheet does that. No document management system does that.
On a typical NZS 3910 project with 150+ Special Conditions, 5 active subcontracts, and weekly correspondence volumes of 30+ letters, the number of active obligations at any point can exceed 200. Tracking these manually requires a dedicated contract administrator working full-time on a single project. When that person is also managing site inspections, attending meetings, and preparing reports, things get missed. The system does not get distracted. It does not forget. It does not go on leave.
Wednesday Afternoon: A Real Decision
The contractor submits a variation claim for additional piling work, citing unforeseen ground conditions. The claim includes a time extension request of 14 working days and a cost claim.
Without the system, the Engineer reviews the claim, checks the drawings, and assesses the merits. That process works. It has worked for decades.
With the system, the Engineer sees the same claim but with additional context surfaced automatically: the original geotechnical report referenced in the contract documents, the specific risk allocation clauses in the Special Conditions, three previous correspondence items where ground conditions were discussed during preconstruction, and the programme impact assessment showing which downstream trades are affected by a 14-day extension.
The Engineer still makes the decision. The system makes sure the decision is made with complete information.
Friday: The Weekly Report
End of week. The system generates a contract health summary: obligations met, obligations approaching deadline, obligations overdue, financial exposure, programme status, and risk flags. The project director reviews a single-page summary that would have taken a contract administrator half a day to compile manually.
An intelligent operating system does not add work. It removes the administrative burden that prevents your best people from doing their actual job: advising clients, making commercial decisions, and managing risk. The system handles the tracking. Your professionals handle the thinking.
Ask your team one question: right now, across all your active contracts, how many obligations are within 5 working days of their deadline? If anyone can answer that question accurately within 30 seconds, you already have a system. If nobody can, you need one.
Provan builds AI-powered operating systems for infrastructure and engineering businesses, covering six domains: Pipeline, Contracts, Projects, People, Finance, and Risk. Every domain described in this article is part of the Provan platform, configured with domain-specific knowledge from 10 years managing projects from $10M to $750M.
This article provides general commentary on construction project management technology. It is not legal, contractual, or financial advice. For specific project management decisions, consult qualified professionals relevant to your situation.
See It in Practice
Book a working session and we will walk through what the system looks like on a project similar to yours. No pitch deck. Real examples.
Book a Working Session →Or email stephen@provan.ai