A leading new energy manufacturer built an AI data platform with MatrixOne Intelligence, enabling end-to-end automated bid document generation and improving bidding decision efficiency by 40%.
Jinpan Technology (SSE: 688676) is a leading Chinese manufacturer of dry-type transformers and energy storage products, supplying new energy, new infrastructure, energy efficiency and rail transit sectors worldwide.
The enterprise faced multi-modal data integration challenges when deploying AI Agent, spanning structured data, semi-structured data, and unstructured data (documents, images, videos).
Bid document preparation involved scattered historical records, high cross-department coordination costs, strict formatting compliance, and inadequate legal clause review.
Built an AI data platform based on MatrixOne Intelligence with three layers: data integration & governance, database & AI services, and application interaction.
Core deliverables include the AI data platform (full-pipeline data architecture) and demonstration AI agents (bid document agent, knowledge base & training assistant, supply chain assistant).
The bid document agent achieves end-to-end automation through AI deep parsing, sensitive information handling, and automated document generation.
Key-information extraction accuracy rose from under 50% to 95%, which largely removes the rework pass that used to follow every read-through.
Locating past-performance evidence across departments dropped from days to 5 minutes.
Bid preparation that took three to seven days now reaches 80% complete in one hour — close to a 50× improvement in overall bid-process efficiency.
With the submission window fixed, a shorter preparation cycle converts directly into how many opportunities the company can pursue at once.
With the AI data platform live, the data and model foundation is settled, several demonstration agents have proven the capability across different business scenarios, and Jinpan moved from a digitalized factory to a smart manufacturing one.
Analysis of business data had been concentrated in the finance department, so every request from the business queued behind finance's schedule.
NL2SQL built on the same AI Factory foundation lets business users query financial and operational data directly in natural language, so analysis is no longer a finance bottleneck.
As a leading manufacturer of electrical equipment, Jinpan signs a large volume of supply and engineering contracts, each reviewed clause by clause by legal.
A contract review assistant reuses the same foundation, comparing incoming contracts against standard clauses and company policy and flagging risks — sharing governance, access control and audit logging with the bid agent and financial analytics.