Boardroom Researcher

Why shouldn't I just paste my transcript into ChatGPT and ask for a summary?

Because ChatGPT gives you a generic summary; this Agent gives you Principal-Level Strategy. We have engineered the system prompt with the specific mental models of a Principal User Researcher. We did the heavy lifting of "prompt engineering" so you don't have to reinvent the wheel every time. It doesn't just "summarize text"; it actively filters for "Binary Trust" moments, behavioral contradictions, and high-risk friction points. You get the strategic depth of a senior hire without having to spend 20 minutes tweaking a prompt for every single interview.

Does this actually write "Strategy" or just summarize what was said?

It is engineered for strategy. We utilize a "Dual-Mode" architecture: The Editor Mode: Forces punchy, 15-word headlines for the slide body so your audience isn't reading paragraphs. The Strategist Mode: Writes deep, nuanced "Director-Level" analysis in the Speaker Notes, capturing the why and the risk, not just the what. Most AI summarizers fail because they try to put everything on the slide. We separated them.

What happens if the AI hallucinates or the data is bad?

The workflow includes a "Diagnostic Safety Net." Before any data touches your slide deck, it passes through a validation logic. If the input is too short (under 50 words) or gibberish, the "Empty Chamber Check" stops the process immediately so you don't waste API credits. If the AI output is malformed, the workflow inserts a specific "⚠️ DIAGNOSTIC ERROR" tag so you know exactly where the break happened, rather than silently failing or printing garbage.

I have my own company branding. Can I still use my own templates?

Absolutely. The system is designed to be "Template Agnostic." You don't need to touch our code to make the slides look like yours. You simply paste our specific "Smart Tags" (like {{user_needs}}) into your existing company master deck. The Engine will automatically map its insights into your layout, fonts, and colors. You control the look; we handle the intelligence.

Can I edit the internal logic or prompts?

The core logic is sealed to guarantee stability. One of the biggest issues with AI workflows is that one small tweak to a prompt can break the downstream JSON formatting. To prevent this, we have "Black Boxed" the reasoning engine. This ensures that the agent always outputs data in the exact structure required to build your slides perfectly, protecting you from breaking the pipeline with accidental edits.

Do I need to be a developer to set this up?

Zero coding required. Since the complex logic is pre-packaged and sealed, setup is strictly "Plug & Play." You don't need to debug JavaScript or understand API schemas. You simply connect your accounts (Google & Anthropic) and start dropping files. We handled the complexity so you can focus on the strategy.

Is my data safe?

Yes. Your data is isolated. Even though the logic is proprietary, the pipeline runs through your secure connection. We do not store, train on, or "peek" at your user transcripts. The Engine acts as a secure pass-through processing layer—it ingests the raw text, synthesizes the strategy, and delivers the final deck directly to your drive, leaving no trace behind.

Can I share this workflow with my entire design team?

This is a Single-User License. Much like a font or a design plugin, the license covers one individual user. While we can't physically stop you from emailing the file to a colleague, doing so violates the license agreement and—more importantly—disables your access to future updates. We frequently patch the "Reasoning Engine" to match new AI models (like Claude 5 or GPT-6); only valid license holders receive these critical upgrades.