How These 3 Founders are building on Open Claw | E2248

What It Covers

Jason Calacanis interviews three founders building on OpenClaw, an open-source agentic AI platform. Presh Dineshkumar demonstrates automated workflows for his wellness apps, Sean Liu shows visual AI integration with Meta Ray-Ban glasses, and Vishnu presents Agent37, a hosted OpenClaw service. Both Sean and Vishnu receive $125k investments live on air.

Key Questions Answered

• •Startup Cost Reduction: Building consumer iOS apps now requires three to four people instead of ten, reducing annual budgets from one million dollars to approximately 300k. Founders can run multiple product experiments simultaneously with agentic tools, testing three to four product theses at once where previously only one was feasible with traditional development approaches and team structures.
• •Automated Customer Support Pipeline: OpenClaw monitors email inboxes for bug reports, searches connected GitHub repositories to identify relevant code sections, diagnoses probable causes ranked by likelihood, and posts detailed analysis to Slack channels. Claude integration then generates code fixes and creates new testing branches, reducing support resolution from hours to minutes without human developer intervention.
• •Multi-Layer Visual AI Architecture: Combining Gemini Live with OpenClaw creates real-time visual understanding through smart glasses. Gemini Live processes camera frames and voice input for perception, then delegates specific tasks to OpenClaw through tool calls. This enables hands-free shopping by identifying products visually, searching Amazon, and adding items to cart through voice commands alone.
• •Automated Content Production Workflow: Setting up 8AM and 6PM cron jobs enables agents to search for industry-relevant research studies, surface findings via Telegram, convert academic papers into digestible blog posts, and prepare social media graphics. This reduces content creation from two hours of manual work to five minutes of review and approval time per piece.
• •Cloud Hosting Economics: Running OpenClaw instances on virtual machines costs approximately one dollar and twenty-five cents monthly for compute resources when properly configured with shared resources. E2-medium instances with four gigabytes RAM provide sufficient performance. Combined with LLM API costs, total monthly operational expenses range from four to ten dollars per user instance.