Yasha Boroumand had been circling AI automation long before it was fashionable. He started programming in middle school, experimented with transformer models in high school, and first tried to build products around URLSlab: a web-change scraper, then a WordPress plugin for technical SEO, content, caching, and image optimization. Both attempts failed, but they left him with domain knowledge around SEO automation and the messy operational work marketers actually need done.

FlowHunt began as a flexible chatbot builder, then evolved into a no-code AI-agent platform for SEO, back-office, support, and sales workflows. The first prototype took about two months. The team began marketing in July 2024, won its first customer in the first month, and landed its first enterprise customer after about six months. The hard part was not just wrapping an LLM; every customer needed a slightly different stack and workflow, so FlowHunt had to adapt across RAG, agentic RAG, AI agents, and changing model capabilities.

The growth lesson is that FlowHunt became its own proof. SEO was the main channel from the start, and Yasha's team used FlowHunt to generate pages, translate content, update and analyze Search Console, and run the workflow end to end. With one copywriter, they produced more than 2,000 pages ranking for 6,000+ keywords and driving about 40,000 organic visits. By the Indie Hackers interview, FlowHunt was doing about $10k/month: roughly $6k from subscriptions and $4k from one-off credits. That split matters: the recurring subscription piece is different from usage-style credit revenue, even when both point to real customer demand.