Hello hello!
I wanted to share a sneak peek at a project I’ve been building. I’ll be sharing a Streamlit Cloud–deployed version soon, but for now the current GitHub repo requires local setup.
Repo:
This is a Streamlit-based AI automation and workflow platform used to run complex, end-to-end automations — especially for ecommerce and small business use cases — while also functioning as a playground for anyone interested in agent building.
Quick Context
Some areas are still evolving and a few things are experimental, but multiple full automation workflows are already working with real platform integrations.
This repo is about sharing a working architecture, not shipping a polished product.
What It’s Capable of Now
- Running full multi-step automation workflows that connect tools and data in sequence
- Generating HTML-rich blogs & emails, scheduling, and even sending them
- Auto-writing and publishing content with automation rules
- Generating product designs and immediately listing them on Shopify via Printify
- Integrating with social, ecommerce, email, analytics, and AI platforms
- Creating multi-clip video ads with music and voiceover
- Uploading to YouTube with metadata & posting automatically
- Browser automation so agents can interact with real web interfaces
Why It’s a Powerful AI Playground Too
Beyond ecommerce use cases, it includes:
- A Replicate playground for experimenting with many models
- A workflow builder for chaining logic, models, and actions
- The ability to generate, convert, and chat with any file type
- A conversational layer that can trigger tools and workflows
Because of this, it also works as a general-purpose automation sandbox for experimenting with AI agents, workflows, and external system integrations.
Why Streamlit Matters Here
Streamlit is used as the control surface to build, inspect, and debug:
- Automation chains
- Agent behaviors
- Integration logic
- Intermediate results at every step
That interactivity has been essential for achieving reliable automation results before abstracting everything into a cleaner interface.
Where This Is Headed
Long-term, this isn’t meant to remain a massive, UI-heavy Streamlit app.
The goal is to extract the underlying automation logic, agents, workflows, integrations, and tools and use them to power a mostly conversational, natural-language-first interface.
Streamlit has been instrumental in shaping the architecture and validating these ideas.
Who This Is For
- Ecommerce founders & small business builders
- Streamlit creators pushing automation beyond dashboards
- AI automation & system-orchestration experimenters
- Anyone who wants a working, extensible sandbox for building automations
Fork It, Remix It, Break It
I actively encourage folks to fork this, simplify it, remix it, or take it in completely new directions.
If you build cool forks, add features, or discover interesting workflows and use cases — I’d genuinely love to see them.
There’s also a GitHub wiki to make the structure, intent, and setup easier to explore.
Final Notes
Happy to answer questions, compare approaches, or even do a brief show-and-tell in an upcoming monthly meeting.
Thanks — excited to see what others might build from it!