Check before you decide.
AI-native product · 0→1 shipped · 1→100 in progress
Try Checkzi Trade · English and 中文
Checkzi is an AI-native decision self-check platform for moments when emotion, urgency, or incomplete information can outrun judgment. It helps people put a decision into words, uses AI within explicit scenario rules to inspect what is driving it, and offers one useful step before acting.
Checkzi Trade is the first and currently only live scenario. It is a pre-trade self-check tool for people with some stock-trading experience who want to avoid rushed decisions when emotion takes over. Before placing a trade, write down your rationale, position size, and emotion. AI turns that context into a structured decision check that helps you see what is driving you, then you decide for yourself whether to act.
Some costly decisions are not caused by a lack of information. They happen in the short distance between knowing one's rules and acting under pressure.
Checkzi is built for that distance: a structured pause that makes the current reasoning visible without pretending that software should take over the final choice.
- Describe the trade in your own words.
- Receive a concise self-check card focused on reasoning, emotion, and plan completeness.
- If useful, add portfolio context for a second check on concentration, pacing, and cash flexibility.
- Save past cards and revisit patterns over time.
Checkzi Trade does not predict prices, recommend securities, or tell a user to buy, sell, or hold.
Future Checkzi scenarios may address other decisions where a short, well-bounded self-check is useful. Each scenario should keep the same core:
- capture the real decision in the user's own words;
- separate facts, assumptions, pressure, and missing information;
- apply rules specific to the scenario before generating explanatory prose;
- offer one proportionate next action instead of an authoritative answer;
- preserve the user's agency and make the product boundary explicit.
Only Trade is live today. Other scenarios are product direction, not shipped capabilities.
- Pause, not predict. Improve decision quality without claiming certainty.
- One useful action first. Keep the first result concise; deeper analysis stays secondary.
- Rules before prose. Deterministic checks handle scenario constraints; the language model explains them in calm, readable language.
- Guardrails are part of the product. Prompting, server-side validation, and interface copy enforce the same boundary.
- Bilingual by design. English and Chinese are supported across the core product journey.
- TanStack Start, React 19, TypeScript, Vite, Tailwind CSS
- Better Auth and Drizzle ORM
- Supabase/PostgreSQL in production
- Pluggable LLM providers with structured output and input safeguards
- Vercel deployment
Checkzi Trade has completed 0→1 and is live. Its 1→100 stage—user validation, growth, and continued product iteration—is in progress. The source code is private. This repository is a public product overview covering the problem, product thinking, current capabilities, and implementation boundaries.
Checkzi Trade is an educational decision-support tool, not investment advice.