Tickerbell

We built this because we needed it ourselves. As we started investing seriously and tried to apply what we were learning from books, the actual workflow sucked: Fundamentals, transcripts, and insider activity were fragmented across different tools. Retail terminals dumped 100+ metrics on us with no indication of what mattered. Screeners were big grids of binary filters—no ranking, assumed expert-level knowledge. There was no fast and easy way to discover stocks that fit a clear thesis plus filters. So we started building the tool we wanted: first a focused ticker page with a decisive subset of metrics, then a screener that ranks by simple signals (e.g. moat + insider buying) instead of forcing dozens of filters. That alone made us much faster at both finding ideas and valuing them. As LLMs improved, we wired them into this stack: per-ticker agents with full structured context (financials, insiders, transcripts, estimates, news) for grounded Q&A and valuation, and a RAG-powered thematic search layer for thesis-first discovery.