By OBBM Network Editorial Staff

Jesse Kennedy and Chris Vasquez

Imagine trying to answer a single customer question while digging through a website, a PDF catalog, three years of reviews, old newsletters and a shipping policy stored in separate folders. The tension of that endless scavenger hunt is familiar to any small‑business owner, and the solution presented by Jesse Kennedy and Chris Vasquez on The Shift AI Show could change the game.

Why a Central Knowledge Base Matters

Small businesses often operate with a patchwork of documentation: a website, a catalog, policy PDFs, and a trove of customer feedback. When a buyer asks, “Do you ship to Canada?” the owner must mentally cross‑reference each source, risking inconsistency or a missed detail. Kennedy notes that “every small business already wrote its own manual. It’s just in six different places.” This fragmentation not only slows response times but also threatens brand consistency.

Enter Gemini Notebook, Google’s rebranded Notebook LM. Unlike generic chatbots, the model only draws from the documents you feed it, eliminating the risk of fabricating answers. As Vasquez explains, “it’s not going to go out and find additional sources and try to fabricate answers based on things that you haven’t already trained it.” For businesses, that means every reply can be traced back to an actual policy or review.

Setting Up the Demo: From Candle Catalog to FAQ

The hosts walk through a live setup for a fictional candle company, Maple and Maine Candle Co., owned by Dana. They upload six source files: a holiday catalog, shipping and return policies, wholesale terms, a batch of customer reviews, and two newsletters. After creating a new notebook on notebook.google, they configure a custom prompt that tells the model, “You are the knowledge assistant for Maple and Maine Candle Co. owned by Dana.” This explicit instruction ensures the AI answers in the brand’s voice.

Once the documents are indexed, Kennedy asks the model to generate a 12‑question FAQ using only the supplied sources. The output includes concise answers in “Dana’s voice” and even flags gaps, listing unanswered topics such as “international Canadian shipping policies” and “specialized packaging improvements.” This dual approach—providing answers while highlighting missing information—offers owners a clear roadmap for updating their documentation.

Beyond Text: Multimedia Possibilities

Gemini Notebook isn’t limited to plain text. The hosts showcase how the same source material can be transformed into a four‑minute podcast script, an explainer video for wholesale buyers, and even infographics. Kennedy pre‑produced these assets to demonstrate the workflow, noting that the AI can generate “video content, audio content, infographics” directly from the notebook’s data.

For businesses with limited marketing resources, this capability could replace costly external agencies. By feeding a single set of documents, owners can produce consistent brand messaging across multiple channels, all while maintaining source verification.

Practical Tips for Small‑Biz Adoption

Both hosts stress that the free version of Gemini Notebook is sufficient for most small enterprises. While paid tiers offer more notebooks and higher source limits, the core functionality—uploading files, setting a custom prompt, and extracting structured output—remains accessible without cost. Vasquez recommends starting with the most frequently asked questions and the most up‑to‑date policies to train the model.

Another key takeaway is the importance of clear prompts. Kennedy advises, “If you have a specific desire, ask for it. Don’t assume that the model will know it.” By explicitly stating the desired format, tone, and source citation requirements, businesses can avoid unexpected outputs and ensure every answer is traceable.

Looking Ahead: AI as the Small‑Biz Brain

The demonstration suggests a future where AI‑driven knowledge bases become the default “brain” for small enterprises. With tools like Gemini Notebook, owners can centralize messy, legacy documentation and instantly generate reliable customer‑facing content. This not only improves efficiency but also builds confidence in brand consistency—a crucial competitive edge.

As the AI landscape evolves, the ability to harness proprietary data without exposing it to broader internet models will be increasingly valuable. Gemini Notebook’s design, which limits knowledge to user‑provided sources, offers a privacy‑first alternative to more open‑ended chatbots, aligning well with the needs of businesses that must protect proprietary policies and product details.

The Shift AI Show’s hands‑on demo underscores that the technology is ready for real‑world use today, not some distant future. Small‑business owners who adopt this approach can expect faster response times, fewer errors, and a stronger, more consistent brand voice across all customer interactions.

The full episode of The Shift – AI Show – Powered by AWeber is available on OBBM Network TV.