🎙️ Voice Agent Demo for Agency Website (Personal Project)
A live voice agent demo built for my agency website to showcase how AI voice assistants can handle real conversations, qualify intent, and respond naturally in real time.
Voice agents are easy to demo badly and hard to demo well. This project was about moving beyond “AI talking” and toward something that actually feels usable for businesses exploring voice automation for sales, support, or intake.
Quick Story: this was my first serious build on Replit.
There were plenty of trial-and-error moments around setup, deployment, edge cases, and stability, but pushing through those challenges gave me a much better understanding of rapid prototyping, iteration, and shipping working AI demos end to end.
20
120
📺 Telegram YouTube Watch Bot (Personal Project)
TGYT-WatchBot is a Telegram bot that monitors YouTube channels and sends instant notifications to Telegram groups or chats when new videos are published.
Instead of checking YouTube feeds manually, updates land exactly where communities already are.
1
19
Skool Community Member Scraper (Personal Project)
A fully automated scraper that extracts structured member data from Skool communities using their Next.js data layer and exports it cleanly to Google Sheets.
Skool does not provide an official way to export community member data.
This workflow allows creators, operators, and growth teams to:
- Analyze member engagement and quality
- Segment users by activity, role, or metadata
- Build CRM, outreach, or analytics pipelines on top of real community data
Instead of manual scraping or brittle scripts, this runs as a repeatable system.
13
97
Recently wrapped a custom analytics dashboard for a client who needed a clearer view of marketing performance without juggling five different tools.
The goal was simple: one place to understand spend, conversion rates, customer value, and channel performance in real time.
Built this as a lightweight internal tool with a clean UI, focusing on clarity over complexity. No bloated features, just the metrics that actually drive decisions.
This kind of work sits at the intersection of funnels, automation, and internal systems. Turning scattered data into something teams can actually use.
If you’re building funnels or running paid traffic and feel blind between tools, this is usually where the biggest wins are.