Sachin Gaikhe·Oct 24Your AI Agent Has Access to Enterprise Data. Who Controls What It Can Do?Your AI Agent Has Access to Enterprise Data. Who Controls What It Can Do?
Sachin Gaikhe·Oct 20Our AI Can Find the Error. Can It Explain Why Production Failed?A payment service starts failing shortly after a deployment.
Sachin Gaikhe·Oct 19From Rules to Transformers: How Enterprise Classification Has ChangedImagine classifying millions of business descriptions into a taxonomy containing 10,000 categories.
Sachin Gaikhe·Oct 11Your LLM Can’t Choose a Category It Never RetrievedIn our previous article, we looked at using embeddings and RAG to classify inputs against a large enterprise taxonomy.A response icon1A response icon1
Sachin Gaikhe·Oct 3AI Costs Are Exploding. How Would You Investigate and Redesign the System?AI systems are easy to make expensive.
Sachin Gaikhe·Sep 26Your LLM Is Not the Classifier: Why Reliable Classification Needs RAGEmbeddings find the possibilities. RAG provides the evidence. The LLM makes the decision.A response icon1A response icon1
Sachin Gaikhe·Aug 1AI Hallucinations: The Hidden Challenge in Enterprise AIOver the past year, one question has come up in almost every enterprise AI discussion I’ve been part of.
Sachin Gaikhe·Jul 25The Art of Chunking: Small Decisions, Big ImpactWhen people talk about improving a RAG application, the conversation usually starts with the language model.
Sachin Gaikhe·Jun 21Why RAG Became the Foundation of Enterprise AIGenerative AI has created enormous excitement.
Sachin Gaikhe·Jun 13AI Will Not Replace Humans, But It Will Change How We WorkOne of the most common discussions around AI is whether it will replace human jobs.