RINDA · GRINDA AI · Product lab · LinkedIn · Email
I'm Hojin Kang, founder and CEO of GRINDA AI in Daejeon, South Korea.
I build AI agents that help smaller teams operate globally. Our flagship product, RINDA, connects overseas buyer discovery, qualification, personalized multilingual outreach, and reply management in one workflow.
The goal is simple: great products should win customers anywhere, regardless of company size.
RINDA helps export-focused B2B teams:
- discover and qualify potential overseas buyers;
- create buyer-specific outreach in multiple languages;
- run follow-up sequences and manage replies in one place;
- keep negotiation and final decisions under human control.
At GRINDA AI, we turn repetitive sales, marketing, and document work into traceable operating systems. I work across product, AI, go-to-market, and the unglamorous last mile between a prototype and a workflow people can trust.
| Project | What it explores |
|---|---|
| claude-rinda-plugin | RINDA workflows for AI coding-agent and MCP interfaces |
| rinda-ai-crm | AI-assisted customer analysis, proposals, and sales-pipeline management |
| rinda-video-ask | Conversational video experiences for product discovery and booking |
| rinda-lead-search | AI-assisted lead and buyer research |
| Saydo | Voice-to-action workflows across software tools |
| VoicePilot | A no-code builder for voice agents |
- Evidence before automation — a useful agent should show why it made a recommendation.
- Human judgment at the boundary — automate repetitive execution, not accountability.
- End-to-end over demo-only — the last mile from signal to action is the product.
- Global by default — language, market context, and local operating constraints belong in the system design.
I'm interested in practical conversations with:
- B2B teams building repeatable overseas sales;
- AI-agent and sales-tech builders;
- global partners working with Korean products and companies.
Explore RINDA · LinkedIn · hogingpt@grinda.ai
Public repositories are a mix of product experiments, prototypes, and open-source learning. Production systems and customer data remain private.