[ TL;DR ]
Explain the margin change, then prepare a reviewable fee request
An engineering project can lose contribution gradually. A client requests another revision, the team carries out an extra analysis and more hours enter the forecast. Understanding the change means bringing together timesheets, appointment terms, client correspondence and design records while the people involved are busy delivering the work.
CSM built a Project Intelligence demonstration workspace for Halden Lowe Building Services Engineers. Its assistant, Aster, explains changes in project contribution, checks additional work against the signed appointment and prepares a fee request for a director to review. Riverside Quarter provides the worked example.
The workspace connects the commercial picture with the records behind it. Aster shows its review steps and links statements to their sources, giving the director an explanation they can check before deciding how to approach the client. The request brings scope, proposed fee, assumptions and exclusions together.
The director retains responsibility for classification and approval. The demonstration shows how preparation can be reduced around that judgement; its financial figures and client agreement are simulated. A live pilot would establish the time returned.
• Less assembling: project hours, appointment terms and client requests are gathered for review.
• Connected evidence: explanations and findings link back to the relevant records.
• Clear responsibility: a person confirms the commercial decisions.
[ The Challenge ]
By the time contribution slips, the explanation is scattered across the project
A project that looked healthy at appointment starts to look thin on the portfolio report. The commercial director needs to understand the movement, but the explanation sits across several records: booked hours, the original scope, client requests and the design team’s response.
Each record answers part of the question. Timesheets show effort. The appointment establishes the agreed service. Correspondence explains what changed, and design records show what the team did. Someone has to bring those parts together before a useful commercial discussion can begin.
Additional work creates another task. A further revision or analysis may need to be assessed against the appointment, described clearly and put to the client as an additional service. Until that happens, the team has delivered work without an agreed additional fee.
Starting late makes the discussion harder. People revisit old messages, reconstruct the sequence and draft an explanation while the project continues. A request based on memory can miss important detail. The opportunity is to prepare a traceable account while the evidence and the client conversation are still current.
[ What CSM Built ]
A workspace that connects project contribution with the evidence behind it
We built a working proof of concept inside the CSM AI Operating System, with Project Intelligence as a workspace branded for Halden Lowe. It brings together project fees, forecast costs, contribution, appointments, client requests, timesheets and design records.
The workspace provides a repeatable process across projects. Each project holds the information needed to investigate its commercial position. The demonstration follows Riverside Quarter, showing how a director can move from a contribution warning to an explanation and a prepared fee request.
Aster supports three connected tasks. It explains why contribution moved using cited records. It compares extra work with the signed appointment. It then drafts an additional fee request containing the scope, proposed amount, assumptions and exclusions for review.
For a client implementation, we would connect the relevant timesheet and accounting systems, document store, email and project records. Access through existing team chat tools would be agreed around the company’s workflow and tested during implementation.
The distinction between preparation and decision remains visible throughout. Aster reads selected sources and prepares the work. A person approves each write and takes responsibility for the commercial decision.
[ Finding the Cause ]
Ask a project question and see how the answer is assembled
The demonstration opens with Maeve, the commercial director, reviewing Riverside Quarter. The project has a €120,000 fee and forecast costs of €98,000, leaving €22,000 contribution. A possible €15,000 additional fee remains outside approved revenue.
Maeve asks Aster to review the project. The assistant shows the records it works through before presenting its explanation. A live step trace makes the review visible, then becomes a receipt of what was checked.
01
Read the project records. Aster works through booked hours, the appointment, client requests and design records. These provide the starting evidence for understanding how the forecast changed.
02
Reconcile the movement. A cost bridge shows forecast cost moving from €80,000 to €98,000. The explanation connects that movement with the scope change and effort recorded against the project.
03
Attach the evidence. Statements carry citations that open the relevant source beside the answer. Maeve can check the record behind a finding while considering the explanation.
04
Keep the review receipt. The steps remain available after the answer is prepared. The director can see what Aster reviewed and assess the explanation against those records.
The result is a starting point for judgement with its evidence close at hand. Maeve can examine the cause of the movement and decide which part needs a commercial response, with less information to gather first.
[ Preparing the Fee Request ]
The Riverside Quarter example turns a contribution question into a reviewable request
With the cause established, Maeve asks Aster to draft an additional fee request for Riverside Quarter. The example follows the same project records through the next stage, keeping the proposed recovery connected to the work behind it.
The client has requested a third revision and a further overheating run. The signed appointment includes two revision rounds and one run. Those records give the director a basis for checking whether the extra work should be treated as an additional service.
01
Bring the evidence forward. Aster works through the evidence cards already connected to the project. The request develops from the reviewed records and the explanation of the changed scope.
02
Describe the additional service. The draft brings the work into a single document, with scope, assumptions and exclusions available for review. Maeve has a prepared account to check against the appointment.
03
Propose the fee. The demonstration request proposes €15,000. That amount remains a draft commercial proposal while the director checks what the document would put to the client.
04
Preserve the cost basis. The €8,400 of delivery effort is already included in forecast cost. The proposed recovery changes the revenue side if agreed; that effort must not enter the forecast a second time.
05
Prepare for the director’s review. The request stays connected to the evidence and financial position. The director checks the classification, amount and assumptions before saving or recording the next step.
This is a simulated commercial example. It demonstrates preparation and financial treatment while leaving approval, submission and agreement with the responsible person.
[ Checking Against the Appointment ]
Establish the scope basis before making the commercial decision
An additional fee request needs a clear explanation of what changed from the agreed service. The appointment is central to that review: it sets the allowances against which the director assesses the client’s request and the engineering work carried out.
In Riverside Quarter, the comparison concerns the third revision and further overheating run. Aster prepares a finding using the appointment and project records, then Maeve confirms the additional-service classification.
01
Identify the agreed allowances. Review the signed appointment’s two revision rounds and single overheating run. These give the reviewer a specific starting point for comparing the additional request with the work already included.
02
Identify the requested change. Bring the client’s third revision and further run into the comparison. Keep the request connected to its source so the director can check the basis of the finding.
03
Prepare a scope finding. Aster describes the work that falls outside the stated allowances and provides the supporting source passages. The director receives a proposed classification to examine.
04
Confirm the classification. Maeve reviews the comparison and confirms the additional service. That confirmation is a named person’s decision, with the relevant evidence available alongside the finding.
05
Use the finding in the fee review. The classification supports the proposed request. The director still decides what to put to the client and how to conduct the commercial discussion.
The workflow prepares a traceable scope comparison. It does not decide contractual entitlement or promise payment. The client conversation and the decision to proceed remain human responsibilities.
[ Review, Submission and Agreement ]
Keep the reviewed request, submission and client agreement as separate decisions
A prepared document still needs an accountable reviewer. Maeve checks the proposed request before saving it, with the amount, owner, scope and assumptions together. The workspace makes the preparation easier while preserving control over what would be put to the client.
The demonstration keeps the stages separate. Saving a draft, recording a submission and recording agreement each represent a different event. A fee enters the approved position only after the agreement is separately confirmed.
01
Is the request ready? Check the amount, owner, scope and assumptions against the project records and appointment comparison. Confirm that the document represents the additional service the director intends to discuss.
02
What has actually happened? Distinguish a prepared request from a submitted request and an agreed fee. Recording submission in the demonstration creates a trail; it does not email the document or recognise additional revenue.
03
Does the financial update reflect agreement? When client agreement is recorded and confirmed, the approved fee changes once. Forecast cost remains unchanged, as do invoices and cash. The €8,400 effort already forecast stays counted once.
These checks let the team follow the request from preparation through the recorded outcome. They also keep a proposed recovery visible without presenting it as approved revenue before the relevant decision has been confirmed.
[ Results ]
A commercial workflow demonstrated from source records to a recorded outcome
The work produced a functioning workspace and a worked example on Riverside Quarter. It shows a contribution review becoming a cited explanation, an appointment comparison and an additional fee request, followed by separately confirmed agreement.
In the simulation, the €15,000 agreed variation moves project contribution from €22,000 to €37,000. Portfolio contribution moves from €208,000 to €223,000. These are demonstration figures, with forecast cost unchanged.
Less evidence to assemble
Aster prepares an explanation and draft from timesheets, appointment terms, correspondence and design records. Directors can review the connected account with the relevant evidence beside it.
Clearer scope and financial checks
The example checks additional work against the appointment and keeps delivery effort already forecast from being counted twice. Pending fees remain separate from approved revenue until agreement is confirmed.
A visible route through the fee decision
The request moves through preparation, review, recorded submission and agreement. Each commercial decision remains with a person, and the source records support the discussion.
These are demonstrated capabilities and intended operational benefits. Submission and agreement are simulated; no real email, invoice or payment is involved. A live pilot would compare the existing process with the workspace, including preparation time, review effort and corrections, to establish net time returned and whether wider rollout is justified.