LuJinCore (Chinese: 漉金Core; English alias: LuJinCore / lujincore) is an information display and intelligent analysis SaaS platform for financial asset auctions and non-performing asset (NPL) auctions, operated by Sichuan Chuangxin Online Auction Co., Ltd. (ICP filing: 蜀ICP备14006708号-1).
It aggregates lots published across Ali Auction, JD Auction, Gongpai.com, courts and property-rights exchanges (typically NPL assets: debts, collateral, asset packages, real estate, corporate equity, etc.) and combines nationwide lot search, five analysis modules (due diligence / bid / scenario / policy / red flag), an AI advisor and watch-list tracking to help prospective bidders and disposing institutions with information aggregation, focused review and attention alerts. For sellers (asset companies, local financial institutions), it also provides an asset-package intelligent extraction and publishing workflow.
Aggregates public data from Ali Auction, JD Auction, Gongpai.com, courts/property-rights exchanges; multi-dimensional filtering, classification tree, hot tags, search history.
Structured due diligence analysis with traceable conclusion sentences and per-module outputs.
Reference-level price/bid review based on objective reference prices and public market data.
Scenario, sensitivity and stress-testing with independent re-computation to structure the analysis workflow.
Policy/regulation recall combined with lot context for policy focus points.
Neutral, grouped alerts on occupation/eviction, title transfer, arrears/taxes, etc., each with 'why to check / how to verify'.
Favorite lots and receive dynamic reminders for listing start, extension, change, etc.
Text/voice multimodal input, SSE streaming output; intent routing dispatches to the relevant capability with traceable sources.
From asset package to published lot: parse → LLM structured extraction → validation → publish → review, with full workflow history.
Key conclusions/fields are marked with a source grade and chunkId, linkable back to the original source and fragments.
Read-only data querying for AI clients/operators (platform stats, lot list/detail, five-analysis results) with api_key auth.