Bollerup-Jensen
Quotation assistant with a data-driven pricing engine
Delivery
Data-driven proposal engine
Azure, dbt, and Streamlit
products
proposal assistant
min. pr. proposal
THE COLLABORATION
Bollerup-Jensen
Bollerup-Jensen is a Danish manufacturing company specializing in silicate-based products, wood care, and cleaning solutions. With a focus on materials technology and product development, the company works to create effective alternatives to traditional chemical solutions with a lower environmental impact. Innovation is a core part of Bollerup-Jensen's work. Through collaborations with research institutions, universities, and industry experts, the company develops products where technical quality, functionality, and environmental considerations go hand in hand. The solutions are used, among other things, in wood care and help meet a growing demand for more responsible materials and production methods.
A manual and complex quotation process with strong potential for automation.
At Bollerup-Jensen, preparing customer quotes was a time-consuming and error-prone process. Even though information about products, prices, and customers already existed in the company's ERP system, each quote required extensive manual calculations and adjustments.
The company saw AI as a potential solution. But the real challenge was getting the existing data and business rules to work together in a more efficient quoting process.
A data-driven quotation assistant that automates pricing and proposals.
Codellent developed a data-driven quotation engine that automates calculations and consolidates the quotation process into a single application. The solution is built on Azure using dbt for data transformation and an intuitive user interface developed in Streamlit.
The quotation engine utilizes data directly from the ERP system, allowing salespeople and management to put together customer-specific quotes with individual discounts, transportation costs, and configurable margins. A targeted, rule-based solution that solves the task without unnecessary AI complexity.
A faster, more accurate, and scalable quotation process.
Bollerup-Jensen can now prepare customer-specific quotes in a matter of minutes. Calculations that previously required manual handling of prices, discounts, and costs are consolidated into an automated process with a consistent data foundation.
This reduces the risk of errors and frees up time for both sales representatives and management. At the same time, the company has achieved a more flexible and scalable quoting process, where individual customer needs can be handled without increasing administrative complexity.
PERSPECTIVE
Bollerup Jensen Group, Working For A Cleaner Future
STARTING POINT
The Ambition
Bollerup-Jensen wanted to make the quoting process faster, more accurate and less dependent on manual calculations. The company saw potential in using AI to automate the work, but first and foremost needed a solution that could make better use of existing data and make it easier for sales staff and management to prepare customer-specific quotes.
The task
Codellent’s task began with identifying the most appropriate technical solution. Rather than introducing AI where it wasn’t needed, the company’s existing ERP data was to be turned into an automated quote engine. This required an end-to-end solution for data integration, transformation and price calculation, as well as a user-friendly application where discounts, transport costs and margins could be adjusted for each individual quote.
IN DEPTH
2025
Case, Bollerup-Jensen
Not all complex problems require artificial intelligence
Bollerup-Jensen's quotation process held significant automation potential. Product information, customer data, prices, and margins already existed in the company's ERP system, but had to be compiled and processed manually every time a new quote was prepared. With individual discounts, transport costs, and various pricing models, even routine quotes became a time-consuming task prone to errors.
Bollerup-Jensen had originally envisioned an LLM-based solution. However, Codellent's assessment was that the challenge could best be solved through data architecture and automated calculation rules. Instead of introducing an LLM, Codellent developed a quotation assistant that uses the company's existing business data as the foundation for accurate and consistent price calculations.
The technical solution is built as a compact data platform on Microsoft Azure. Here, dbt is used to transform data from the ERP system, allowing information on customers, products, prices, and margins to be integrated into the quotation engine's calculations. On top of this data foundation, Codellent developed an application in Streamlit where sales representatives and management can compile customer-specific quotes, adjust discounts, and factor in transport costs.
This has transformed a previously manual and complex process into a far more efficient workflow. Quotes that previously required extensive calculations and manual handling can now be prepared in just a few minutes. At the same time, price calculations are consolidated into a single solution, making them more consistent and easier to maintain as products, prices, and business needs evolve.
THE COLLABORATION
Bollerup-Jensen
OVERVIEW
A custom-built proposal engine that automates complex price calculations and streamlines customer-specific quotes
Codellent
Use data and knowledge to create better offers
Delivery
Data-driven proposal engine
Industry
Production
Area
Use data and knowledge to create better offers
Effect
Faster time to market, fewer errors, better scalability
(03)
PROJECTS



