Stay updated with our thoughts on AI, automation, and digital transformation
In many user-facing applications, such as documentation search or user manuals, the goal is to retrieve relevant text while matching user intent. Traditional RAG approaches combine text indexing through embeddings with semantic search, but this creates a fundamental problem.
Read article
In the race to build intelligent machines, large language models have emerged as sophisticated pattern-matching systems rather than true reasoners.
Read article
Your AI specialists will explain that AI typically involves preparing data by cleaning and processing. But what if we could bypass these steps entirely?
Read article
When we first considered developing a system using LLMs to analyze communications, we explored three different approaches to goal achievement.
Read article
While traditional IQ tests are designed to measure human intelligence across areas, this approach fails when applied to AI systems.
Read articleThe real value of AI lies in finding effective solutions through methods that would be impractical or impossible using deterministic approaches alone.
Read articleHow our experience in banking software led us to build a new kind of AI - one that combines language understanding with logical reasoning.
Read article