The network for creativity
Join 1.25M professional creatives like you
Connect with clients, get discovered, and run your business 100% commission-free
Creatives on Contra have earned over $150M and we are just getting started
Back to feedPost
🚀 FlexUI AI — Generative UI Dashboard Designer
FlexUI AI is a Generative AI-powered dashboard designer that converts natural-language requirements into interactive dashboards.
Users can describe what they need in plain English, for example:
“Create a sales dashboard showing revenue, orders, monthly trends, regional sales, and top products.”
FlexUI AI uses Google Gemini to understand the request and generate a structured JSON dashboard specification, which is dynamically rendered into an interactive Streamlit UI.
Key Features
Natural language → Dashboard
AI-generated JSON UI schema
Dynamic KPI cards
Interactive Plotly charts
Bar, Line, Pie & Scatter charts
Dynamic data tables
JSON schema preview
Tech Stack
Python · Streamlit · Google Gemini API · Plotly · Pandas · JSON
Concepts
Generative UI · Structured Output · Prompt Engineering · LLM Integration · Dynamic UI Rendering · Data Visualization
This project demonstrates how Generative AI can go beyond text generation and dynamically drive user interfaces and visualizations.
The network for creativity
Join 1.25M professional creatives like you
Connect with clients, get discovered, and run your business 100% commission-free
Creatives on Contra have earned over $150M and we are just getting started
Related posts
Built an AI-powered Social Media Post Analyzer that analyzes social media content and generates structured insights using LLMs.
The application uses LangChain and Pydantic to process the input and return structured results through an interactive Streamlit interface.
Built with: Python, LangChain, Pydantic, Hugging Face, Streamlit, and LangSmith.
Pydantic catches shape errors, but a plausible wrong insight can still pass. I'd keep a small set of posts with expected labels in LangSmith and rerun it after prompt changes. Are you tracking that kind of drift?
From an idea to a working Figma plugin. 🎨🚀
I've been working on something exciting: Acquire Color, a Figma plugin that extracts colors from images and turns them into structured color systems.
The idea was simple, but bringing it to life has been a learning experience.
I used AI to help me build it, and along the way, I experienced a few ups and downs.
The challenges:
Getting AI to understand my exact vision.
Refining the UI when the initial results weren't what I wanted.
Figuring out how to structure primitive and semantic color systems.
Debugging, iterating, and rethinking certain parts of the experience.
The wins:
Turning my idea into a working prototype.
Exploring Figma plugin development.
Learning to collaborate with AI more intentionally.
Developing a clearer vision of what the product should become.
One thing I've learned is that building with AI doesn't mean everything happens automatically.
You still need to think through the problem, make decisions, guide the process, and refine the output.
I'm still improving Acquire Color, but seeing the idea take shape is a win in itself.
A little progress is still progress. And I'm enjoying the process of becoming a better designer and builder.
What has your experience been like building with AI?
#Contra #BuildInPublic #ProductDesign #Figma #AI
End-to-end automation of an auto glass business working with insurance companies. Insurer emails and documents are read automatically, the case is opened, the right glass is matched by VIN and compatibility, supplier stock and prices are compared, installation is booked, a payment link is issued and the client is kept informed at every step. People only make decisions. Live with real orders every day.
Reading insurer emails and documents automatically and matching glass by VIN really shrinks the manual data entry step. Issuing a payment link and booking installation keeps the process moving without extra clicks.