Founded in 2003, uEngine Solutions is a technology-driven software company with its own core technologies in Business Process Management (BPM), AI, and microservice architecture (MSA). Its process management technology is used by major Korean manufacturers and research institutes including Samsung Electro-Mechanics, LG Display, LG Energy Solution, LG Chem, Hyundai, Kia, Hanwha, and the Korea Institute of Industrial Technology. So far, AI has remained a personal tool — drafting documents, summarizing text. Process GPT makes it the actual executor of organizational work. When material arrives late, who checks what, in what order, and reports to whom? That sequence usually lives only in one person's head. Describe it in plain language and it becomes an executable process, with AI agents querying ERP, WMS and SCM systems directly to carry out the work. People step in only where cost and accountability are at stake, such as approving an alternate purchase order — a mixed workforce in which people and AI agents operate as one team. Manufacturing does not need a plausible answer; it needs the same answer every time. Process GPT solidifies a verified execution path into code and reproduces it identically from the second run onward with zero LLM inference, and every agent decision remains traceable for quality and audit review. Because systems are linked through an ontology knowledge map rather than consolidated into one database, it can be adopted without replacing existing ERP or MES systems.

Process GPT — AI Agent Platform for Manufacturing Business Processes AI agents execute cross-departmental work spanning order intake, production planning, material procurement, quality and shipping. · Supply-chain delivery-risk response agents — a single customs hold handled end to end as it cascades into procurement, inventory, purchasing and customer response, across ERP, WMS, customs and CRM (live demo at booth) · Automated reverse-engineering of work standards and quality/safety documents — executable BPMN processes extracted from PDF and image files · Natural-language process generation — describe the work and get a BPMN process, input forms, human/AI role assignment and approval steps · Deterministic Replay — verified execution paths solidified into code, reproduced identically from the second run with zero LLM inference · MCP Hub — enterprise system integration (ERP, WMS, SCM) with tool-level access control and instant rollback · DMN rule-based decision control — Human-in-the-Loop approval wherever cost is incurred · Ontology Studio — knowledge graph linking inconsistent part numbers across systems, aggregating from source systems without data migration · Execution Profiler and Lead-time Breakdown — isolating human judgment, agent execution and queue time to expose bottlenecks Applicable processes — delivery delay response, alternate sourcing review, 4M change management, nonconformance (NCR) handling, customer claim response, supplier evaluation, trade documentation, internal policy review