AI and software for contractors
We do construction software development for contractors, installers and engineering firms, and apply AI in construction where documents pile up: tenders, specifications, quality dossiers and subcontractor paperwork. Site managers and estimators keep the decisions, with better information in front of them.
Where the work gets stuck.
Tenders hide risk in hundreds of pages
A tender package holds specifications, drawings, memos of information and contract terms. Estimators have days to find the requirements, deviations and penalty clauses that decide whether a bid makes money.
Quality evidence is collected too late
Under the Wkb the contractor must show the building meets the rules at completion. Photos, checklists and test reports are scattered over phones, apps and email, and the dossier gets assembled in the final weeks.
Site and office run on different data
Planning, costs and progress sit in King or Syntess, drawings in Autodesk Construction Cloud or Procore and site notes in WhatsApp, so the project picture is always out of date.
How we improve construction.
Each one is scoped around your systems and rules, and each one keeps a person in charge of the decisions that matter.
Tender and specification analysis
Tender documents are searched and summarised into requirements, deviations, risks and questions for the memo of information, each linked to its page. Estimators and the tender manager decide what to price and what to ask.
Wkb completion dossier assembly
Photos, checklists and inspection reports are sorted by building element and risk from the quality plan, and missing evidence is flagged while the work is still accessible. The site manager and quality assurer sign off the dossier.
Offline site inspection app
A mobile app for inspections, snags and progress photos that works without signal in basements and on remote sites and syncs to your project systems. Findings are assigned to subcontractors with a deadline.
Subcontractor invoice and contract matching
Subcontractor invoices are read and checked against the contract, approved quantities and extra work orders, including chain liability requirements such as the G-account share. Project controllers approve or dispute each invoice.
BIM model questions for engineers
An MCP server that lets an assistant query IFC models for quantities, element properties and clashes, so engineers can ask questions without opening modelling software. Answers link back to the model element.
Project cost and progress data
ERP costs, planning and site progress combined into one data model per project, refreshed daily. Project managers see margin and schedule risk before the monthly review.
Built around the rules.
What we design for from the first week. Your legal and compliance people keep the final word.
Wkb (Wet kwaliteitsborging voor het bouwen)
Quality evidence is tied to the risk assessment and the quality assurer's plan, and the completion dossier keeps a traceable record of who recorded and approved what.
Omgevingswet and the Bbl
Requirements from the Besluit bouwwerken leven and permit conditions are referenced in checks and dossiers so they can be verified against the current rules.
Chain liability (Wet ketenaansprakelijkheid)
Invoice checks record G-account splits, subcontractor documentation and staffing evidence, so the finance team can show what was verified for each payment.
GDPR for site workers and photos
Site photos and access records can identify people. We limit what is stored, set retention per project and keep data in the EU.
Works with what you run.
If a system has an API, a database, an export or an inbox, we can build on it. These are the ones we meet most.
- Autodesk Construction Cloud
- Procore
- Trimble Connect
- BIM and IFC models
- Snagstream
- King ERP
- Syntess Atrium
- Microsoft Dynamics 365
- Relatics
Where to start.
Tender risk scan for one business unit
Tender analysis is document work with a clear deadline and a clear judge: your estimators. We build a search and summary tool for one type of tender and test it on bids you have already priced.
Talk it throughWhat it includes
- Ingestion of specifications, drawings lists and contract terms
- Requirement and risk extraction with page references
- Question list for the memo of information
- Evaluation against recent tenders you know well
Guides.
AI readiness checklist: what to answer before your first agent goes live
The questions to answer before your first AI agent handles real work, from process and data to approvals, monitoring and ownership.
10 min read
Build vs buy software: a decision guide for operations and product teams
A practical guide to choosing between off-the-shelf and custom software, with a cost comparison, a decision matrix and a checklist for your team.
10 min read
Further reading.
How to build a production RAG chatbot: a practical guide
What it takes to turn a promising RAG experiment into a chatbot people trust: ingestion, hybrid search, citations, permissions and evals.
5 min read
How to evaluate RAG: retrieval metrics, faithfulness and golden sets
How to measure a RAG system properly: separate retrieval from answers, check faithfulness claim by claim and build a golden set you can trust.
4 min read
RAG vs fine-tuning: which does your business actually need?
RAG gives a model your knowledge at answer time; fine-tuning shapes its behaviour. How to choose, when to combine them and what each costs.
4 min read
Common questions.
Yes. Both have APIs for documents, issues, RFIs and forms. We read from them for search and reporting and write back issues or forms where it saves double entry.
Yes. Inspections, photos and signatures are stored on the device and sync when the connection returns, with conflict handling if two people edit the same item.
Current models read text, tables and title blocks on drawings reasonably well and struggle with precise measurements. For quantities and geometry we work from IFC models, and a person checks anything that feeds a price.
Probably not. We usually connect to the app you use, such as Snagstream, and add the missing pieces: automatic sorting of evidence, gap checks and the link to your ERP and planning.
A tender risk scan for one type of tender generally takes weeks, depending on how many past tenders we can test against. We agree a timeline once we have seen your documents.