FAQ
Frequently asked questions
Clear answers about wallet credit, usage, subscriptions, and how Tycoon charges for work.
What is the single most important thing to get right when hiring an AI employee?
The job description. Everything else follows from it. A clear job description tells you what model and platform to pick (based on the required skills), how to configure the agent (which tools, what memory), how to onboard it (what SOPs and reference cases), and how to review it (what success looks like). Operators who skip this step end up with agents that are technically capable but organizationally useless. The time spent writing one-page job descriptions pays back within the first month.
Should I use one big AI platform or assemble my own stack?
For most solo founders, a platform wins in the first 6-12 months. Tycoon, Polsia, and similar systems handle the plumbing (memory, tool use, workflows, dashboards) so you can focus on configuration and oversight. Assembling your own stack (raw OpenAI / Anthropic APIs + custom orchestration + custom memory) is a valid choice once you have very specific needs, but it usually adds 20-40 hours per week of infrastructure work that does not move the business. Start on a platform. Move off it only when the platform is constraining something measurable.
How much should I pay AI employees?
Most productive AI employee configurations cost $20-200 per month in model inference and platform fees, depending on volume and model choice. A full 5-10 agent team typically runs $500-3,000 per month at reasonable scale. This is not meaningful money relative to the output they produce; do not try to optimize model costs before you have optimized their quality. Once you have a team shipping good work, you can revisit model choices to capture savings.
How do I fire an AI employee?
Identify the repeated failure pattern. Try three configuration changes: expand memory, change model, tighten scope. If performance does not recover after 2-4 weeks, replace the agent. Replacement usually means one of: a different model on the same platform, a different platform entirely, or breaking the role into two narrower roles. Firing is not emotional with agents the way it is with humans, but the discipline of doing it matters — it keeps your team from slowly degrading under the weight of underperformers.
What does this look like in Tycoon specifically?
Tycoon models AI employees as first-class roles. You hire an AI CEO, AI CMO, AI CTO, AI COO, AI CFO, and functional reports from a role catalog. Each role ships with a default job description, skill set, memory scaffold, and workflow library — you customize from there rather than starting from scratch. Onboarding, review, and replacement are supported as workflows, not as ad-hoc chat sessions. The goal is to make hiring AI employees feel like the most important HR decisions at a normal company, because that is what they are becoming.