TrendLens

Companies can collect endless customer reviews and competitor research — but turning it into a real marketing decision is still hard. TrendLens is an AI-powered marketing opportunity discovery and validation tool that closes that gap for any industry. A company uploads its customer feedback and competitor messaging, and TrendLens follows a four-step workflow: Discover the recurring customer needs, Connect them to what competitors are (and aren't) advertising, Validate each opportunity against the actual evidence, and Recommend a testable A/B campaign with a success metric.

The open-weight Gemma 4 model (served via the Google Gemini API) does the language work — extracting needs, writing the grounded recommendation, and drafting the experiment — while deterministic Python does all the scoring, so every number is transparent and traceable to real records. Its signature feature: add a competitor message and the opportunity score updates live, letting you falsify a recommendation instead of blindly trusting the AI. Unlike a dashboard or a scraper, TrendLens never asserts a number it didn't measure — it recommends a test, and shows its evidence.

Built with: Python, Streamlit, Gemma 4 (open-weight) via Google Gemini API, Plotly, Pydantic. Open-source (MIT).

Built With

• claude
• gemma