Frequently Asked Questions
Everything you need to know about LexCore's accuracy, privacy, and workflows.
CitationGate is our trust mechanism. When the multi-agent AI identifies a contract risk, it records the precise character start and end offsets of the proving clause. If a quote cannot be verified against the source, the finding is rejected — so the AI cannot surface a claim it can't trace to your document. It reduces hallucination risk dramatically; it does not replace a lawyer's review, which is why Human Gates keep you in control.
No. LexCore enforces a strict data-privacy policy. Your documents and annotations are saved inside isolated, encrypted tenant databases and are never used to train public models.
LexCore is pre-seeded with rules for major common-law jurisdictions, with a specific focus on the Nigerian legal framework (including CAMA execution requirements and NDPA-aware handling), alongside standard UK and US-DE company configurations.
When you run an AI redline, LexCore generates suggested edits and rationales. When you click download, our document mechanics construct a native Microsoft Word (`.docx`) file where all changes are pre-anchored as standard tracked changes. You can open and edit it directly in Word.
Yes. For enterprise teams requiring maximum data sovereignty, we offer a self-hosted package allowing complete deployment of the AI orchestrator, vector memory databases, and local models (using vLLM / Ollama) behind your own private cloud infrastructure.
You can upload a contract or select a sample on our scanner page. The scanner runs a single review round and displays 2 findings with exact citation matches to demonstrate our platform's accuracy. The remaining findings are gated behind a free account registration.