Pradeep Bansal·Jun 14Once Upon a Time I Used My Brain… Now I Use AI TokensPart 1: How AI, Screens, and Modern Life Are Quietly Eroding Your Brain
Pradeep Bansal·May 12Abstract AI Agent: Your AI Agent may be doing too muchThe case for thin agents with smart tools — and why your 8,000-word agent prompt is a liability.
Pradeep Bansal·Dec 15, 2025Advanced Text2SQL: Moving Beyond Keyword Matching in Enterprise EnvironmentsIn the last post, we concluded that to solve text-to-SQL, we had to stop being the student and become the teacher who rewrites the exam…
Pradeep Bansal·Sep 11, 2025The Perfect Query Paradox: Why 99% recall in text-to-SQL is not enoughAiming for perfection may not be ideal thing to do in real life, but this is what an ideal text-to-SQL system demands. In fact, working on…
Pradeep Bansal·Sep 1, 2025Lost in Translation: Why Text-to-SQL Breaks Down in Enterprise DatabasesIn this first post of our three-part series, we explore why translating natural language questions into SQL queries is fundamentally…
Pradeep Bansal·Jun 11, 2024LLM Hallucination Detection: Background with latest techniquesOne major hurdle in adopting AI for real life scenario is its inability to be accurate every time. There was a sudden “hope” with the…
Pradeep Bansal·Jun 10, 2024LLM Hallucination Detection: Can LLM-Generated Knowledge Graphs Be Trusted?An LLM response can be hallucinated which means it can be factually incorrect or inconsistent w.r.t. the reference document. For eg. while…
Pradeep Bansal·Oct 1, 2023Fine-Tuning a pre-trained LLM with unlabelled datasetA generic LLM or an LLM pre-trained on the dataset different from the domain of the final task has to be fine-tuned to the task specific…A response icon1A response icon1
Pradeep Bansal·Sep 27, 2023How do you use BERT as a classifier without fine-tuning?BERT, Bidirectional Encoder Representations from Transformers, is the deep learning model quite famous for generating embeddings that are…
Pradeep Bansal·Aug 7, 2023Common questions while using the Hugging Face’s transformers library.Instead of writing one complete blog on topics like summarization or text generation, I thought of taking most important topics involved…