The European glass industry works with high climate and quality targets and must simultaneously withstand global competition. The young company Lumeso, Vienna, shows how this can work and strengthens European manufacturers where rapid results can be achieved in process optimisation.
The start-up’s AI-based software minimises the time from order acceptance to delivery and is now being used successfully by several large glass manufacturers. Georg Katzlinger-Söllradl, CEO of Lumeso, explained, “We won’t be able to achieve ‘cheap production’ in Central Europe. Manufacturers from countries with lower labour costs, minimal environmental standards and low energy prices are better at that. But what glass processors in Europe can definitely do better is customer proximity, individuality and speed. Now that many manufacturers have already automated their production, the automation of upstream sales processes is becoming the most important lever.”
Many manufacturers in the European glass industry have invested in the automation of production processes in recent years. The challenges and opportunities are now shifting to upstream processes such as quotation submission, order entry, technical clarification and work preparation. Lumeso’s AI-based solution therefore automates manual processes in internal sales, which until now have mostly been carried out manually. In particular, order data exchange is greatly simplified, thus increasing productivity in the back office.
One of Lumeso’s early adopting customers is Olympic Glass, in the UK. Adam Roda, Operations Director at the British glass manufacturer, reported, “Lumeso’s AI-based solution brings targeted automation to our back office, reads enquiries from emails and PDFs, links items to our master data and thus brings verified data records into the ERP. We see this as a great opportunity to significantly reduce processing times from enquiry to start of production, thereby drastically improving the customer experience. Lumeso’s support and response times have been exceptional so far, so the software’s performance is already stable after some early fine-tuning.”
Olympic Glass is already using the next version (Lumeso 2.0), which displays a multi-step process on one screen instead of three and is even more user-friendly. Identical product structures can also be more easily assigned to multiple order items, and different delivery addresses are recognised.
AI explained simply: What LLMs can do
Artificial Intelligence, and as a sub-discipline, Large Language Models (LLMs), have quickly established themselves as an essential component of solutions for automating processes, especially in connection with unstructured data. LLMs act as intelligent ‘reading and comprehension programmes’ that receive emails, PDFs, tables or drawings, recognise important information such as dimensions, variants and quantities, summarise content and generate suggestions for quotations and orders.
The difference to classic, rule-based software is that LLMs can also handle unstructured documents. Employees can focus on the final check, while the AI takes care of the time-consuming preparatory tasks. This means that orders are better checked and often enter the system in minutes instead of days, reducing the time to production.