Challenge
The construction industry is highly dependent on the efficacy of a firm’s tender process. Dealing with issues such as risk management, detailed specifications, project complexity and multiple stakeholders, a firm’s response to a tender document literally gives them a competitive edge – or not. Like a lot of other firms, BAM, the leading market player in area development, residential and commercial construction, technology, infrastructure, and mobility, driven by the ambition to build a more sustainable Netherlands, faced significant challenges with managing tenders.
Tender submissions require handling vast amounts of information, including unstructured data, different legislation that needs to be considered, various preference documents, project risks and more. These processes are extremely labour-intensive and time-consuming, which made it difficult for BAM to run an efficient, accurate and quality focused tender process.
BAM wanted to streamline its process to ensure it could efficiently process and verify large volumes of tender documentation and improve the accuracy and speed of information extraction. It also wanted to be able to automate repetitive and manual tasks to reduce the workload on teams and enhance the classification of project risks for better decision-making.
Solution
BAM and Valcon are working together to develop a suite of AI-driven tools designed to streamline the whole tender process and achieve all its objectives.
The first is the development of ‘Tender Assist’ which has a user-friendly interface where users can interact with the tender documentation. The app aggregates all the tender files, extracting essential information and linking back to source documents. It allows users to quickly validate information by providing references to original sources. Using the built in LLM functionality, users can query the data using natural language.
Using generative AI, the solution automates processes that previously required significant manual effort. It uses a multi agent approach which allows the users to interrogate the tender documents using the factual agent and helps to write the tender by examining old tender documents and using a creative GenAI model. It also allows users to interrogate tender documents, which reduces the need for manual research.
Benefits
The implementation of these AI-driven tools has provided several key benefits for BAM:
- Productivity gain: the solution has significantly improved the efficiency and productivity of the team by reducing the time spent on processing tender documents, which allows teams to focus on strategic tasks.
- Enhanced accuracy and validation: all extracted information can be quickly verified, which reduces the risk of errors in tender submissions.
- Reduced manual workload: the use of AI and generative AI eliminates labour-intensive processes, helping users navigate to the sections they need to read much faster than they could do themselves.
- Scalability and flexibility: the integration of the latest open AI models into the Mendix app offers flexibility for users to query and interact with data in real-time.
BAM’s Tender Assistant, which is now in production, is part of a larger programme to harness the power of AI and automation in construction processes. This partnership with Valcon has set a foundation for future innovations, ensuring BAM stays ahead in an increasingly competitive and digital construction industry.
