Systemic

Ever wish you could just see how everything connects? Powered by Perplexity AI, Systemic generates insightful systems maps in Miro, revealing hidden relationships & the bigger picture.

Comment

Inspiration

I've always been fascinated by how complex systems work. The challenge is that building systems maps can be very time consuming and difficult. I wanted to use Perplexity AI's research power to instantly create research-based systems maps, making complex topics intuitive and easy to understand. That’s why I built Systemic - to turn prompts into clear systems maps in Miro.

What it does

Systemic is a Miro app that takes your text prompt, uses Perplexity AI (via an n8n workflow) to research and identify key elements and relationships, and then draws them as an interactive systems map directly on your Miro board. It helps you quickly see the bigger picture of any topic.

How I built it

Systemic was built solo during the Perplexity Hackathon. Frontend: A Miro app (HTML, CSS, JS/React with Vite), deployed on Vercel. Backend: A cloud-hosted n8n workflow orchestrates calls to the Perplexity Sonar API, processes the data, and sends it to Miro. AI: Perplexity AI's Sonar API does the heavy lifting of research and structuring the information.

Challenges I ran into

Prompt Engineering: Iterating on the system prompt for Perplexity to get consistently good, structured JSON output for mapping was the biggest challenge. Miro Visualization: Translating AI output into a visually coherent map on Miro using the SDK took some fine-tuning. Time: As a solo developer in a hackathon, focusing scope was key.

Accomplishments that I'm proud of

Building a fully functional, end-to-end AI application as a solo developer within the hackathon timeframe. Developing an app which builds on an area that I'm passionate about, and which I feel can (with some continued iterations!) truly positively impact people. Creating a clean, usable Miro app interface that's live and testable on Vercel.

What I learned

Perplexity AI is incredibly powerful for structured data output if you craft your prompts well. n8n is fantastic for rapid backend development and API orchestration, especially for a solo dev. The direct visual impact of a systems map greatly enhances understanding complex information.

What's next for Systemic

I plan to make Systemic more powerful with better research capabilities and references: Iterative Research: Enable multi-step Perplexity queries for deeper dives into sub-topics. Source Linking: Have Perplexity provide justifications/sources for map elements. AI-Powered Analysis: Use Perplexity to analyze the generated map for system archetypes and feedback loops. Interactive Exploration: Allow users to click nodes to trigger further Perplexity research on specific elements. My goal is for Systemic to be the go-to tool for AI-assisted systems thinking.

Built With

Share this project:

Updates