{"id":3883,"date":"2025-11-13T16:23:02","date_gmt":"2025-11-13T14:23:02","guid":{"rendered":"https:\/\/www.puurdata.nl\/?p=3883"},"modified":"2025-11-17T10:20:44","modified_gmt":"2025-11-17T08:20:44","slug":"how-to-find-what-you-are-looking-for-in-unstructured-data","status":"publish","type":"post","link":"https:\/\/puurdata.nl\/us\/zo-vind-je-wel-wat-je-zoekt-in-ongestructureerde-data\/","title":{"rendered":"Smart with AI \u2013 How to find what you're looking for in unstructured data"},"content":{"rendered":"
<\/div><\/div><\/div>
\n\t
\n\t\t
\n\t\t\t
\n\t\t\t\t
<\/div> <\/div>
\n\t
\n\t\t
\n\t\t
\n\t\t\t

Smart with AI - How to find what you're looking for in unstructured data<\/h1>\n
\n\t
\n\t\t\t<\/div>\n<\/div>\n\n\n\n\n\t\t<\/div> \n\t<\/div>\n\t<\/div> \n\n\t
\n\t\t
\n\t\t
\n\t\t\t\n\t\t<\/div> \n\t<\/div>\n\t<\/div> \n<\/div><\/div>\n\t\t\t<\/div> \n\t\t<\/div>\n\t<\/div> \n<\/div><\/div>\n\t\t
<\/div><\/div><\/div>
\n\t
\n\t\t
\n\t\t\t
\n\t\t\t\t
<\/div> <\/div>
\n\t
\n\t\t
\n\t\t
\n\t\t\t\n
\n\t
\n\t\t

by Francesca Brzoskowski<\/em><\/p>\n

Imagine this: a colleague in operations wants to know how your company resolved a customs delay in Germany last year. She knows someone has picked it up and that emails were circulating about it. But as soon as she types "Germany customs 2024" in the shared folder, 47 emails and 30 PDF reports appear, most of which have nothing to do with her question.<\/strong><\/p>\n

She asks three colleagues for help. One sends a report, another adds someone from logistics. An hour later, she has eleven browser tabs open, but still no clear answer.<\/p>\n

This is a problem in almost every organization. Not because people are disorganized, but because they lack the resources to search effectively.<\/p>\n

Traditional search works only to a limited extent<\/h3>\n

Many organizations still rely on traditional search methods. These primarily search for exact word combinations: in emails, files, folders, or subfolders.<\/p>\n

If you type "onboarding issues," you'll only see documents containing exactly those words. A typo or a missing word will cause important results to disappear.<\/p>\n

Search smarter with semantic and vector search<\/h3>\n

New search technologies work differently. Whether you're using a modern search engine or an MCP server connected to your apps, these systems understand the meaning behind words.<\/p>\n

Text is converted into vectors, or numbers that represent the meaning of words. This allows a search engine to understand that a query about onboarding issues could also refer to documents about customer implementation, setup challenges, or delays with new accounts.<\/p>\n

An MCP server goes one step further. It searches all your data simultaneously\u2014emails, documents, PDFs, shared drives\u2014and automatically determines where and how to search.<\/p>\n

\"\"<\/p>\n

A practical example: from hours to seconds<\/h3>\n

A manufacturing company had fifteen years of maintenance reports, all saved as PDFs. Technicians knew that certain machines had recurring problems, but finding similar incidents usually meant reading through hundreds of reports.<\/p>\n

After implementing semantic search with Elasticsearch, an engineer could simply ask, \u201cWhat typically causes heating in hydraulic systems?\u201d<\/p>\n

Within six seconds, he received six relevant incidents, spanning different years, locations, and formats, but all related in content. Manual searching would have taken hours, if anyone even bothered to do it.<\/p>\n

When do you choose which approach?<\/h3>\n

The best approach varies from organization to organization. Here are the key considerations.<\/p>\n

Semantic search engine (e.g. Elasticsearch)<\/strong>
\nWhen building your own semantic search engine, the data must first be prepared: slicing, embedding, vectorizing, designing indexes, and testing them with real queries. It's a lot of work, but it delivers predictable and easily verifiable search results. Ideal if you want concise, precise answers.<\/p>\n

MCP servers<\/strong>
\nWith MCP servers, you can skip that preparation. They work directly with what you already have: emails, tools, documents, shared folders. However, you do need to ensure proper authentication, permissions, and monitoring. MCP servers are faster to implement, but more difficult to manage without clear operational agreements.<\/p>\n

The way forward for organizations<\/h3>\n

Technology is no longer the limiting factor. Embedding models are accessible, and MCP servers are maturing rapidly. What matters most is where your organization stands:<\/p>\n