AI transforms Supplier Quotations into Structured Business Data
Elkedjan
AI Turns Administration into Customer-Focused Sales Time
Background
Elkedjan is Sweden’s largest nationwide chain of electrical installers, coordinating purchasing and supplier interactions across a highly distributed network. Every month, hundreds of quotation requests are sent to suppliers across the market. The responses arrive via email, often as lengthy PDFs and attachments containing pricing, product details, delivery terms, and technical information.
Large amounts of critical business information existed inside unstructured documents, requiring extensive manual handling before it could be used operationally. Data had to be reviewed, interpreted, and manually entered into internal ERP systems before quotes could be approved and processed further. As quote volumes increased, so did the administrative burden. The process became time-consuming, difficult to scale, and heavily dependent on manual work.
Together with Esatto and Ecometrix, Elkedjan initiated a project to explore how AI-driven document decoding could automate and streamline this workflow.
The Assignment
The objective was to create a scalable solution capable of transforming supplier quotations from unstructured PDFs into structured, system-ready data. The ambition was not simply to digitise documents, but to create a more intelligent, scalable, and data-driven operational workflow.
The set-up focused on:
Automated decoding of supplier quotation documents
Extraction of relevant commercial and product information
Structuring outputs into formats compatible with ERP workflows
Creating a scalable foundation for future workflow automation
The solution was built using Zeno, Ecometrix's AI-powered document processing platform. By combining advanced document understanding, AI-powered extraction, and enterprise-grade cloud infrastructure, the platform can interpret supplier quotations regardless of format and transform them into structured, system-ready data.
This enables Elkedjan to process quotations faster, reduce manual administration, and create a foundation for future workflow automation.
AI as an Operational Accelerator
We quickly demonstrated the practical value of AI-driven document processing. Instead of manually reviewing and entering information from supplier quotes, AI models were able to extract and structure relevant data automatically. Information previously locked inside PDFs became immediately accessible and operationally usable.
The workflow combines OCR-based document understanding with LLM-driven interpretation and deterministic mapping logic, enabling the system to handle varying supplier formats while still delivering structured outputs compatible with downstream ERP processes.
At the same time, the project reinforced an important insight about applied AI: The greatest value is not only the automation itself, but the ability to create scalable, repeatable workflows around data that was previously difficult to utilise.
By combining AI-based extraction with human validation, Elkedjan could improve both efficiency and data quality while continuously refining the models over time.
This creates a powerful operational foundation for future automation initiatives.
“ Before we started the project, our goal was to make the quotation and sales process more efficient. The work has shown us how AI can turn a traditionally manual and time-consuming process into a much more scalable and efficient workflow. What impressed us most was not only the speed of the document handling, but the ability to structure and utilise information that previously was locked inside PDFs and emails. This creates entirely new opportunities for automation and operational insights going forward. Thanks to the AI agent, we can now free up time and use it for more valuable sales work with our customers.’’
— Mats Liljegren, VD Elkedjan AB
Built for Enterprise Requirements
Because supplier quotations often contain commercially sensitive information, security was a key requirement from the outset. The solution is hosted in European cloud environments, supports GDPR-aligned processing, and includes encryption, access controls, and full auditability.
This enables Elkedjan to benefit from AI-driven automation while maintaining the security and governance standards expected of a modern enterprise solution.
" What made this project successful was the smooth cooperation with Elkedjan and the clear business objective. The goal was never to automate documents for the sake of automation—it was to free people from administrative work so they could spend more time creating value for customers. That's where AI delivers its greatest impact.”
— Fredric Belin, Esatto
How We Work with AI-Driven Automation Today
Every AI project contributes to improving how we design and implement future workflows. Three important learnings from this project have become central in how we approach AI-driven automation today:
1. AI creates value when connected to real operational workflows: The strongest impact comes when AI is integrated directly into existing business processes, not treated as an isolated technical feature.
2. Structured outputs are the key to scalability: The real business value emerges when unstructured documents are transformed into structured, reusable data that can flow into existing systems and reporting processes.
3. Human validation accelerates long-term accuracy: Combining automation with human feedback loops creates more reliable workflows while continuously improving AI performance over time.
The Results: Freeing time for more sales
For Elkedjan, the project established a significantly more scalable and efficient approach to quotation handling.
More time for valuable sales work: By automating administrative tasks, we free up time for better and more meaningful sales work with our customers.
Reduced manual administration: AI-driven extraction reduces the need for repetitive manual data entry and document handling.
Faster access to operational data: Supplier quotations become immediately available for review, analysis, and downstream processes.
Improved consistency and quality: Standardised extraction improves consistency across supplier documentation and reduces the risk of manual errors.
Foundation for workflow automation: The project creates a pathway toward future automation of the entire quote-to-order process, from incoming email to ERP registration.
Better utilisation of business data: Structured quotation data provides improved insight into supplier activity, pricing, and operational performance.
Want to know more?
Daniel Nilsson
Managing Partner Stockholm
