TransparencyWins
Case studies
AI Reduces Geotechnical Document Analysis from 3 Weeks to 3 Hours

Case study

AI Reduces Geotechnical Document Analysis from 3 Weeks to 3 Hours

ConstructionUnited States

Case study summary

A US-based geotechnical company implemented an AI document-processing system to automate analysis of multi-page PDFs containing borehole logs, tables, diagrams and soil-layer drawings. The solution used LLMs and VLMs to identify document structure, separate borehole logs, extract soil-layer data, store structured results in a database and generate client-formatted Excel spreadsheets. A proof of concept tested the models on real client data against expert-prepared evaluation sets. Document analysis time fell from about three weeks to around three hours, while quality was reported as comparable to human experts.
View original case study

TransparencyWins ecosystem context

This case study was contributed by VM.PL Software House, a software engineering partner represented in the TransparencyWins ecosystem. Case studies make delivered project experience, project context and reported outcomes visible to tech buyers. Review this contribution together with available client testimonials, partner insights, certifications and other company signals when evaluating relevance for a sourcing need.