ClaudeLab

Use case · Construction & building services

Tenders in construction and building services

In construction and building services, reviewing tenders is the most expensive unpaid part of sales. A bill of quantities has hundreds of items, most awards lead to no order - and still someone has to read every set of documents to know that.

Highest value per individual case

What is different here

What it actually turns on in construction and building services.

01

The bill of quantities is structured - but not uniform

GAEB files can be split up mechanically, PDF bundles and scans cannot. In practice both arrive, often in the same case, and every issuing body numbers its items differently.

02

Eligibility criteria decide before the price does

References, certificates, accreditations, minimum turnover: anyone who does not meet them need not cost the job at all. These criteria sit scattered through the tender documents and are regularly overlooked.

03

Deadlines are hard

A bidder question has a cut-off date, and so does submission. What arrives late is not evaluated - however good it was. A pre-selection that costs two days therefore sometimes costs the participation.

Systems we connect here

  • GAEB files
  • Tender platforms
  • Tendering software
  • Nevaris
  • ORCA
  • In-house service catalogues

If yours is not listed, that is not an exclusion. Where no interface exists, we build one - that is the most demanding part and it is our trade.

The rule that only applies here

In public procurement, the pre-selection decides the return, not the costing. Every hour spent on a tender you were going to lose anyway is missing from the one you could have won.

The process

Where a person is needed - and where they no longer are.

  1. ClaudeLab from here

    Step 01Person

    Documents

    Intake

  2. Step 02ClaudeLab

    Review

    Checking

  3. Step 03ClaudeLab

    Matching

    Deciding

The difference

The same process, once as it is today and once with us.

Step 01 · today

The documents arrive as a PDF bundle, sometimes with hundreds of items.

Step 01 · with us

The bundle is read out in full, even at hundreds of items.

Step 02 · today

Someone skims them and decides whether costing is worth it.

Step 02 · with us

The item list is held against your range of services, with exclusion and eligibility criteria alongside.

Step 03 · today

If it is costed, items are matched by hand against your own range of services.

Step 03 · with us

You get an assessment with reasons and decide whether to take part.

What we build

Component 01
The bill of quantities is broken into items and held against your range of services.
Component 02
Exclusion and eligibility criteria are pulled out and checked.
Component 03
The result is an assessment with reasons, not a yes-no traffic light.

Degree of autonomy

The evaluation runs independently; the decision to take part is made by a person. A tender wrongly dropped is never seen again.

Where it does not fit

Where an award turns on references and relationships, an evaluation helps little - there it is not the document that decides.

Frequently asked

Is this worth it in construction and building services?

In public procurement, the pre-selection decides the return, not the costing. Every hour spent on a tender you were going to lose anyway is missing from the one you could have won. Whether it pays for you depends on your volume - the audit establishes that.

We use a different system from the ones listed. Does it still work?

Yes. The ones listed are those we most often meet in construction and building services - a list, not a precondition. Where no interface exists, we build one.

What do we have to provide?

Your range of services in structured form - a list is enough, it only has to be complete. Ten to twenty past tenders together with the decision you took at the time. A contact in costing for the questions of the first weeks. Nothing more is needed to start.

How will we know it works?

Across 20 real tenders the "cost it / drop it" recommendation agrees with yours in at least 85 %. The item list is read out in full - verified by spot check against the PDF. No bid is submitted without a person looking at it.

How long does implementation take?

5 to 8 weeks - the most demanding of the eight cases.

Does the AI decide on its own?

The evaluation runs independently; the decision to take part is made by a person. A tender wrongly dropped is never seen again.

When is this not worth it?

Where an award turns on references and relationships, an evaluation helps little - there it is not the document that decides.

The same process, in general

This page shows the fit in construction and building services. The process itself is sector-independent - what matters is whether it repeats.

Tenders in general

Tell us where your time goes.

In the audit we record the process, work out what is worth automating, and you get a plan with fixed prices - credited against the build if you go ahead.