Few presentations at the 2026 Value Investing Seminar took a purely practical turn, but Blagoj Janevski’s session did. Janevski is a self-employed value investor based in North Macedonia, a solopreneur by his own LinkedIn description rather than the principal of a named fund or firm. He started his career as a software developer and internet marketer before moving into long-term value investing. He now runs his own SaaS business and reinvests its profits and dividends into public equities.
His talk was a hands-on tour of Google’s NotebookLM, a source-grounded AI research assistant he considers the most useful tool available today for serious financial analysis. His argument: querying large volumes of financial filings is where general-purpose large language models like Claude or ChatGPT fall short in ways that matter, and NotebookLM’s architecture addresses those shortcomings directly.
What NotebookLM Is and Why It Is Different
NotebookLM, Janevski explained, is not a general chatbot. It is a research assistant that operates exclusively on documents you supply. You upload PDFs, websites, YouTube videos, audio files, or spreadsheets, and the tool answers questions only from those sources, citing the exact passage behind each answer. The design eliminates most of the hallucination risk that makes general-purpose AI unreliable for financial work. The tool runs on Google’s Gemini model and started as a small internal research project before expanding into broader use.
Login required to continue reading.
Setup a free account to get access to this article (no credit card required).
View Full Article
