← Back to ProjectsCase Detail Jiedanba · OPC Trading Matchmaki
Jiedanba · OPC Trading Matchmaking Platform
R&DEnterpriseLive

Jiedanba · OPC Trading Matchmaking Platform

Institutional-Grade Escrow · AI Smart Matching · Milestone Delivery Guarantee

ReactTypeScriptNode.jsPostgreSQLDrizzle ORMpnpm monorepoViteExpressTailwind CSSe签宝 V3AI AgentRadix UIView Project
screenshot-1
2条
A/B Dual-Channel Independent Signing Link
全链路
Requirements → Quotation → Contract → Delivery → Settlement Closed Loop
3类角色
Issuer / OPC / Platform Operations Tri-Portal Workbench

Project Summary

Jiedanba is a transactional matchmaking platform that connects institutional clients with high-end OPCs (Super Individuals). It covers an end-to-end, closed-loop process including requirement posting, AI-assisted quoting, electronic contracts (eSign V3), fund escrow, milestone delivery, and ticket support. Dual contract channels (A: Platform ↔ Client; B: Platform ↔ OPC) are signed independently, with the entire contract lifecycle managed by the platform.

Background & Challenge

Traditional outsourcing involves extremely high trust costs: clients struggle to verify the actual delivery capabilities of OPCs, payment security for OPCs is unassured, contract signing relies on offline workflows, delivery processes lack structured management, and dispute resolution lacks traceable evidence.

My Role

The core dilemma of the OPC economy is the trust issue in high-value transactions between strangers. The platform's value lies in building a sufficiently robust trust infrastructure—fund escrow ensures payment security, electronic contracts (supporting dual-mode signing for both enterprises and individuals) guarantee legal validity, and credit tiers and rating systems enable capability screening. Embedding AI Agents into quotation and contract draft generation can significantly reduce operational headcount, serving as a key lever for the platform's evolution toward lean operations.

AI Integration

Multi-turn ReAct Agents drive OPC quote recommendations and contract draft generation, with cross-turn state accumulated via `historyMessages`; e-Signbao V3 dual paths—the standard path (template + variable filling + automated platform seal → counterparty signing URL) and non-standard path (direct PDF upload + keyword coordinate positioning)—run in parallel, supporting conditional enterprise/individual org signers; a pnpm monorepo three-package architecture (jiedanba frontend / api-server backend / db shared layer) leverages Drizzle ORM schema-driven database migrations to ensure multi-environment consistency.

Implementation Process

1
Phase 1: Core Platform Framework Setup (User System, Requirement Posting, Dual-Portal Workbench, Role & Permission Management); Phase 2: Transaction Matching & Contract Management (Bidding & Awarding, Dual-Channel E-Contracts, Fund Escrow, Milestone Delivery); Phase 3: AI Agent Integration & E-Signature Closed Loop (ReAct Quoting Assistant, End-to-End eSignbao V3, Contract Template Configuration System)

Key Deliverables

Jiedanba OPC Transaction Matching Platform (Live in Production);
A/B Dual-Channel Electronic Contract Signing System (eSignbao V3 Integration);
AI Smart Quotation & Contract Draft Generation Agent (Multi-turn ReAct);
Platform Admin Backend (Contract Templates / Placeholders / Role & Permission Configuration);
End-to-End Dual-Side Workbench for Order Issuers / OPCs

Reflection

The eSignbao V1→V3 upgrade represents the most significant architectural refactoring milestone: the API shifted from direct file upload mode to template generation mode, requiring seamless integration across three components—the frontend template selector, the backend `sourceField` variable parsing pipeline, and the conditional org signer—where a failure in any single link breaks the signing workflow. Enterprise signer qualification strongly depends on the integrity of settlement account data, necessitating a graceful fallback prior to initiation. The `sourceField` mapping mechanism for contract placeholders (in `table.column` format) is a crucial design for long-term system scalability—adding new fields only requires maintaining the `placeholder_defs` table, with no code changes required.

Media Coverage

Interested in this technology?

Whether you are a potential partner, recruiter, or technical evaluator — start a conversation anytime.

Get in Touch