Blogs

How will McKinsey and Accenture create value in the age of AI agents? We walk through the four strategies CB Insights identifies — integrating the technology stack, turning proprietary data into an asset, productizing services, and human-AI collaboration — and show how each takes concrete form in uEngine solutions such as Process GPT and Ontology Studio, in our four-stage transformation roadmap, and why open source is the final piece.

When a customs delay hits, how do you connect procurement, inventory, purchasing, and customer support? Walk through a demo that uses Ontology Studio and Process GPT to analyze the impact of the delay and handle everything from alternative-order approval to the customer reply.

See which tools AI agents use, where they use them, and under what conditions they fail. Learn how Process GPT's tool analytics compares usage and failure rates by department, process, and execution engine to pinpoint what to improve.

Running multiple MCP servers means managing permissions, versions, and call history alongside connection settings. See how to register servers in Process GPT's MCP Hub, verify connections, and review per-tool permissions and operational status.

Total elapsed time alone rarely explains why work is delayed. Learn how Process GPT separates human review and approval, agent execution, and queue wait time to find bottlenecks, and how to correctly read the work-time axis and estimated values.

Reviewing work done by AI agents requires their inputs, outputs, and tool-call records. See how the Process GPT Execution Profiler shows processing time and token cost per task and lets you trace down to individual calls.

The declaration that "BPM is dead" was a category error from the start. What died was the heavy, rigid 2000s-era BPMS tool, not BPM as the management discipline of designing, measuring, and improving processes. In the era of the triple workforce, where people, AI agents, and robots work side by side, work allocation is no longer a one-time design decision but an operational variable that must be continuously re-optimized, and process is its coordinate system. This piece covers the view that 'graph engineering', now emerging in the engineering community, is BPM's execution layer while BPM is its missing management layer; how to redefine EA as infrastructure for speed of change rather than control; and how uEngine6 BPM and Process GPT turn this discipline into product.

A report on lean AI-native companies, analyzing Midjourney, Cursor, Chatbase, Cal AI, and others on the Lean AI Native Companies Leaderboard (over $5M ARR, under 50 employees, less than 5 years old). Starting with the revenue-per-employee leaderboard of Telegram, Midjourney, and Anysphere, it examines the structure of a business model that "sells outcomes, not labor", the management principle of "automate before you hire", and the traps hidden behind ARR and headcount figures. It also maps these principles to Process GPT features and AX Transformation Consulting Training, then closes with the areas Korean companies can apply and an 8-step execution framework for delivering 100-person results with a team of 10.

A write-up of the special keynote CEO Jinyoung Jang delivered at the 2026 Open Source Developer Challenge orientation. It covers the 23-year growth story of an open-source company that began with the grand prize at the 2003 competition, an open-source business strategy that monetizes 'operational risk' through the Trust Funnel, and AI-era technology trends such as the end of coding and the rule of the spec, coding intelligence doubling every 70 days, and the collapse of the 1:8 golden ratio. It closes with the role shift to product engineer, the 30-minute iteration cycle, three survival rules, uEngine's 20-year answer leading to Process GPT and Robo Architect, and a developer manifesto for the AI era. The presentation slides are available as a PDF download.

Examines the design challenges of cloud-native transformation: service boundaries, data ownership, and transactions. Using the Gov24 case and the challenges at the design, implementation, operations, and deployment stages, it lays out a path through the transition.

Real competitiveness comes not from AI technology itself but from becoming an 'AI-Native Enterprise' — an organization that redesigns judgment and execution from the ground up on the premise that AI exists. This is uEngine Solutions' blueprint: from simple adoption to full redesign, and how Ontologic Platform and Process GPT turn that picture into real products.

The questions raised by Palantir's rapid growth and arrival in Korea — why ontology has become essential infrastructure for agentic AI now, and what a platform must have to qualify, examined through the design of Ontology Studio and Ontologic Platform.

The self-improvement mechanism of 'loop engineering', the hot topic in the AI agent industry, is a perfect mirror image of BPM's philosophy of continuous process improvement (CPI). Follow the journey of these two worlds converging into 'Agentic BPM', and explore the Process GPT features that have already turned that blueprint into a product.

How far has it come, is it ready for my project, what changes if I adopt it, and which solution should I choose? Answers to the questions of client organizations, development firms, and PMOs, backed by data and case studies.

This post walks through the design and analysis process of the first phase of the Gov24 portal cloud-native transformation project carried out by uEngine Solutions, and explains in technical depth the changes brought by adopting microservices architecture (MSA).

In the business world, traditional Business Process Management (BPM) is facing the most significant shift since its inception. The global BPM market was valued at roughly $18.9 billion in 2024 and is projected to surge to $65-76 billion by 2037.

The dilemma of brilliant but forgetful AI - If you have used conversational AI like ChatGPT, Claude, or Gemini, you have probably had this experience: an AI that gave strikingly accurate and useful answers at first starts forgetting what was said earlier or losing consistency as the conversation grows longer and more complex.

When migrating complex PL/SQL code to Java or refactoring large legacy codebases, we often run into unexpected bugs. Finding and fixing every problem by hand is tedious. But what if the code could analyze its own logs, find the cause, and fix itself?

Cohesion describes how closely the elements within a module are related. A highly cohesive module has a clear, focused purpose. Coupling refers to the degree of interdependence between different modules. Low coupling increases module independence and greatly improves the flexibility and scalability of a system.

In software design, CQRS (Command Query Responsibility Segregation) is a useful technique for meeting the demands of complex applications. In particular, separating read and write operations improves performance and makes the system easier to manage. In this article, we look at how to implement CQRS with Spring Boot.

In a monolithic architecture, all customer-related data is often consolidated into a single "Customer" table. Because this table holds data from many domains (e.g., sales, marketing, reservations), it is known as a "God Table".

Microservices Architecture: Understanding It Through the Factory and Warehouse Analogy
This training material explains the concepts of microservices architecture through the analogy of factories and warehouses. The main topics are as follows.

Today we discuss one of the biggest challenges in microservices architecture: "balancing data consistency and performance". In particular, we look at how the 'aggregate' concept from domain-driven design (DDD) and the 'eventual consistency' pattern solve this problem.

Cohesion and coupling are fundamental concepts in software design. Today we use everyday objects as examples to make these concepts easy to understand, then apply them to object-oriented programming and microservices architecture.

This article explains how to apply domain-driven design (DDD) principles to solve common problems when using JPA/Hibernate. It introduces a range of issues you may encounter with JPA and Hibernate and presents approaches for resolving them.

Once upon a time, the village where the three little pigs lived was a peaceful and safe place. Each had a house of their own and lived happily. Through this story, we can see three different approaches.

We look at how to modify entities to handle, through polymorphism, a case where employees are classified as contract or full-time and the tax calculation logic differs only for contract employees.

Writing an Oracle Stored Procedure to calculate payroll for all employees can involve the following steps: calculating base salary, overtime pay, deductions for unpaid leave taken, and tax withholding.