SignalCut：将AI搜索可见性缺口转化为视频营销活动的智能工具

SignalCut is an innovative web application that analyzes brands' visibility gaps in AI search, generates evidence-based marketing strategies, and creates Hera video materials. It helps early-stage brands gain a competitive edge in the AI answer engine era by bridging the gap between visibility analysis and actionable marketing execution.

Traditional SEO focuses on rankings in search result pages (SERP), but AI answer engines like ChatGPT and Perplexity directly generate answers instead of listing web links. Even well-ranked brands may be ignored in AI responses. Early brands face a vicious cycle: less visibility in AI answers → less exposure → harder to grow, as AI tends to favor well-known enterprises due to their higher frequency in training data.

SignalCut was developed at the Big Berlin Hackathon. Its core idea is to transform AI search visibility analysis into executable marketing actions, forming a complete workflow from insight to execution by integrating multiple AI technologies and platforms into a commercially valuable application.

SignalCut's workflow forms a closed loop:

1. Input Info: Brand name/website, competitors, target audience, Peec AI visibility context data.
2. Identify Gaps: Analyze Peec AI data to find high-value prompt gaps (important to the audience but where the brand underperforms).
3. Gather Evidence: Use Tavily API for real-time retrieval (search volume, user discussions, industry reports, competitor strategies).
4. Generate Strategy: Use Google Gemini API to create core messages, audience segmentation, content themes, and channel suggestions.
5. Create Videos: Convert strategies to video ideas via Hera API, generating optimized variants for different audiences or channels.

Frontend: React 19 (better performance), TypeScript (type safety), Vite (fast build), Lucide icons. Backend/Deployment: Cloudflare Pages (hosting/edge computing), Pages Functions (serverless API handling), Wrangler (CLI tool). AI Integrations: Google Gemini (strategy/creative/prompt optimization), Tavily (real-time evidence), Hera (video generation), Peec AI (visibility data input). Security: API keys stored server-side (via Cloudflare Pages Secrets) to avoid browser exposure; evidence retrieval ensures fact-based strategies.

Brand: Attio (a CRM software) Competitors: Salesforce and HubSpot Identified Prompt Gap: "Best CRM for agencies needing custom onboarding workflows" SignalCut's actions: Analyzed visibility gaps → identified the prompt gap → gathered evidence via Tavily → generated a full campaign strategy → created video variants → produced MP4 via Hera → provided exportable campaign packages (Markdown or JSON format).

Implications:

• Closed loop from monitoring to action (reduces insight-to-action friction).
• AI-driven content production (automates execution, letting marketers focus on strategy).
• Combinatorial innovation (combines specialized APIs for powerful apps).
• Evidence-based marketing (data-backed strategies for credibility).

Limitations: Dependence on external APIs (availability/cost), AI video quality not yet professional-level, strategy depth may lack human expertise.

Future: Support more content formats (text, podcasts), integrate more data sources, add A/B testing, develop industry-specific templates.

SignalCut is an open-source project. Deployment steps:

1. Clone the GitHub repository.
2. Install Node.js dependencies.
3. Configure API keys (Tavily, Gemini, Hera).
4. Deploy to Cloudflare Pages using Wrangler.

For developers: It serves as a learning case for modern web apps (Vite/React), Cloudflare Functions, multi-AI integrations, secure key management, and end-to-end workflows.