When Linara Bozieva, founder of Ravenopous, lost her job at eBay after over 10 years with the company, she faced a difficult period. She had recently moved from Switzerland to the United States with her family, and the tech industry was going through major layoffs.
Instead of searching endlessly for another corporate role, the 39-year-old decided to build something herself, reported Business Insider. A few months after the layoff in 2024, Bozieva launched Ravenopous, a marketing agency based in San Jose. But unlike traditional agencies, much of the work inside her company comes from artificial intelligence systems she built on her own.
Currently, her business runs on a three-layer AI workflow powered by 27 custom AI agents. The system handles tasks such as ad creation, campaign analysis, traffic generation and customer conversion. Bozieva said the entire setup costs her less than $1,000 a month, reported Business Insider.
“It felt worth it to start building something small, at least for a couple of years, to support my family and me,” she told Business Insider.
Before starting her company, Bozieva spent 11 years at eBay, where she worked in analytics. She said that experience helped her understand how to structure the AI system and build the right guidelines for the agents.
At first, she experimented with tools such as ChatGPT and Midjourney. Later, she became interested in AI agents that could operate in closed-loop systems with minimal human involvement.
She initially built her workflow on Google’s Antigravity platform but later shifted to Claude Code after running into token limits with Gemini Pro. “I love Claude, but I have started hitting token limits there as well,” she told Business Insider.
How does Linara Bozieva’s AI system work?
Bozieva told Business Insider that the workflow operates through three separate layers – directives, orchestration and execution. The directives layer defines each agent’s role, skills and knowledge base. The orchestration layer acts like the brain of the system and decides which agents should work on specific tasks.
Six agents operate inside this orchestration layer. They include a market researcher agent, a data analyst agent, a creative director agent, a finance agent, a legal agent and an orchestrator agent that sends tasks to the execution layer.
Once tasks move into execution, different agents handle technical work, traffic generation and customer conversion. Bozieva told Business Insider three agents focus on technical setup, 10 handle awareness and traffic campaigns, while five focus on turning customers into revenue.
The workflow also includes scripts that automate repetitive tasks. Some scripts even pull customer pain points directly from Reddit discussions to sharpen marketing strategies.
Bozieva said the AI system manages ads, tracks performance, improves creatives and delivers daily reports. Clients receive weekly reports and monthly review calls.
The setup relies on several paid AI tools. She pays for Claude Code, Codex and ChatGPT subscriptions, reported Business Insider. She also uses services such as HeyGen and ElevenLabs, along with APIs that connect multiple AI models together.
Human judgement required?
Even though AI performs much of the work, Bozieva said human oversight remains essential. “AI can process a transcript of a client call, but it cannot fully read the room or identify where the client seems most nervous,” she told Business Insider.
She added that strategy decisions still require human judgment, especially when there is limited data or when the AI system suggests multiple paths forward.
Bozieva also believes expertise matters more than ever in an AI-driven world. According to her, AI can help automate systems, but people still need deep knowledge of a subject to guide the technology correctly.
“I think all repetitive tasks can and will be done by AI in the future,” she said. “However, the expertise needed for specific departments or domains will still be with humans acting as VPs over AI architecture,” she told Business Insider.
The founder tested the workflow on 14 client profiles before officially relying on it for paid business. Some tests came through free projects for friends, while others came from her own internal experiments.
Currently, she has five active clients using the AI-driven system. Once clients complete onboarding, Bozieva spends around two hours a week managing each account, reported Business Insider.
She believes the setup could eventually support 20 to 25 clients without requiring a large team. Instead of hiring traditional employees, she plans to recruit operators who supervise AI agents and guide overall strategy.
“Building something on my own has been incredibly fulfilling and promising,” she said. “It changed not just how I work, but what I believe one person can build,” she told Business Insider.