An AI-driven investigation workspace. You give it a subject — a person, entity, or claim — and it does the research, lays out the findings as a board you can navigate, mark up, and act on.
You enter a subject and a bit of context. PlanChart runs a multi-round investigation using Claude (an AI model by Anthropic) plus live web search. Findings are rendered as a board of connected cards — facts, sources, claims, and the relationships between them. You can drill deeper on any card, mark what matters, snapshot evidence before it disappears, and re-run a query from a different angle.
Not a Consumer Reporting Agency. PlanChart is not registered, does not operate, and is not intended as a Consumer Reporting Agency under the US Fair Credit Reporting Act (FCRA) or equivalent regulations elsewhere. Outputs are informal pre-vet material and must not be used as the basis for adverse employment, credit, housing, or insurance decisions. If you need a formal background check, use a CRA. PlanChart is for the work you'd otherwise do in a browser tab and a notes app — only better.
Not a court-grade evidence system. Snapshots and exports are evidence preservation for your investigation, not chain-of-custody, not sworn, not legally admissible without independent verification.
Not a stalking, harassment, or targeting tool. See Terms and our internal principles. Investigations matching patterns of stalking, doxxing, targeting minors, or targeting people for membership in a protected class are refused.
Not anonymous. We keep audit logs of sensitive actions (see Privacy). If you need anonymous research, this is not the product.
We keep what's needed to run the service honestly: the subjects you submit, the rounds we ran, the sources we found, your annotations, and audit logs of sensitive actions. We do not sell user data. We do not serve ads (no AdSense, no third-party trackers). For the full breakdown, see Privacy.
Each investigation costs roughly $1–$3 in AI tokens plus amortized infrastructure. PlanChart is currently free to use, supported by a tip jar — if it's useful to you, you can buy a coffee to keep it running. A metered API and paid tiers are planned; pricing will be transparent and published when those launch.
This is early. The engine works, the board works, the policies above are real. Some advertised features (live voice / Telegram mode, external API, snapshot/archive) are in build. If something is broken or missing, email below.
Built by Daniel as a serious tool for the kind of investigative work people do badly because the tooling is bad. Currently early — Daniel is the first user, friends and family are next, the rest comes after.
Questions, feedback, or takedown requests: hello@planchart.ai