Challenge
VAM Entreprenør, a leading Danish construction company, faced a highly manual and fragmented tender process. Project teams spent countless hours collecting documents, formatting material lists, requesting sub-vendor prices, and preparing the Bill of Quantity for their final offer. This slowed down their ability to respond quickly and consistently to tender opportunities.
The Problem
- Heavy manual file handling across many formats
- Time-consuming preparation of Bill of Quantity
- Sub-vendor correspondence done manually for every tender
- Offers received in different formats, difficult to compare
- Final offer assembly required significant senior oversight
- No simple way to compare new tenders to historical ones
Solution
We developed TenderCore, an AI-powered tender assistant that automates and structures the entire tender flow. Instead of spending time on administration, VAM’s project team can now focus on crafting the strongest possible final offer.
- Document Intelligence: AI reads and understands all tender documents, regardless of format, and transforms them into a structured Bill of Quantity foundation.
- AI-Driven Workflow: The system automatically manages the tender flow end-to-end.
- Automated Sub-Vendor Handling: TenderCore prepares emails, sends out Bill of Quantity requests, and processes incoming offers.
- Smart Offer Translation: Sub-vendor prices are normalized and converted into structured data usable directly in the Bill of Quantity.
- Bill of Quantity Assembly: The full Bill of Quantity can be compiled with internal resource libraries and vendor inputs.
- Tender Assistance: AI helps the team review all documents, compare to similar historical tenders, and optimize the final offer.
Technology Stack
- GPT-4 and domain-adapted LLMs
- Custom NLP models for Danish construction tender terminology
- Python backend with automated document processing
- Vector database for historical tender and offer retrieval
- Integrations for sub-vendor communication and file ingestion
- Power BI for tender insights and comparison analytics
Results
As TenderCore is being rolled out, VAM is already seeing transformative impact:
- Up to 75% faster tender preparation by removing manual file work
- No manual formatting, AI structures all tender documents automatically
- Consistent Bill of Quantity creation across project teams
- Sub-vendor flow automated, reducing administrative load significantly
- Higher-quality final offers created with more time for analysis and fine-tuning
- Instant access to past tenders, improving pricing strategy and accuracy
Key Features
Smart Document Processing
AI reads all tender documents (PDFs, spreadsheets, drawings, and more) and converts them into a clean, structured Bill of Quantity starting point.
Automated Sub-Vendor Management
TenderCore prepares sub-vendor emails, sends requests, receives their offers, and transforms all incoming pricing data into a unified format.
Bill of Quantity Assembly & Optimization
Project teams can build the final Bill of Quantity using internal resource libraries and AI-translated vendor inputs, ensuring consistency and transparency.
Intelligent Tender Assistance
The assistant provides an overview of all tender documents, highlights key elements, and compares the tender with similar historical projects.
Business Impact
- Tender teams save significant time through file automation and workflow orchestration
- Sub-vendor coordination becomes structured, consistent, and fully traceable
- VAM can create stronger final offers due to more time for strategic work
- Improved accuracy in Bill of Quantity and pricing through AI-driven normalization
- Historical comparisons help improve competitiveness and quality
- Ultimately enabling VAM to deliver better tenders with less effort
TenderCore is now expanding to support even more of the tender and project planning phases.
