Author

DeWu Technology

A platform for sharing and discussing tech knowledge, guiding you toward the cloud of technology.

429

Articles

Likes

2.9k

Views

Comments

Latest from DeWu Technology

100 recent articles max

DeWu Technology

Oct 14, 2026 · Backend Development

Taming LLMs with Deterministic Workflows: Dynamic Rule Engine for Asset Acceptance

Dewu Technology shares a production system that uses LLMs for asset acceptance verification across dozens of expense types with frequently changing rules, replacing hard-coded logic with a configurable rule engine, deterministic workflow orchestration, multi-model voting, and a closed-loop quality system with backtesting gates and human feedback.

LLMProduction Engineeringasset acceptance

0 likes · 43 min read

Taming LLMs with Deterministic Workflows: Dynamic Rule Engine for Asset Acceptance

DeWu Technology

Sep 30, 2026 · Mobile Development

How Dewu Delivers 60FPS 3D Spatial Photos on Mobile with Gaussian Splatting

Dewu's engineering team details their end-to-end pipeline for turning single photos into interactive 3D spatial images using 3D Gaussian Splatting, covering model generation, perceptual quality evaluation, 81.5% model compression, and mobile GPU optimizations that achieve stable 60FPS rendering on iOS and Android.

3D Gaussian splattingDISTSGPU radix sort

0 likes · 17 min read

How Dewu Delivers 60FPS 3D Spatial Photos on Mobile with Gaussian Splatting

DeWu Technology

Sep 2, 2026 · R&D Management

AI-Native Delivery Evolution: Harness Framework for Controllable, Auditable Engineering

This article details Dewu's 'Delivery Harness' system—Version Contract, Execution Boundary, Evidence Gate, and Repair Loop—that enables a solo developer to manage AI-assisted full-stack delivery across multiple runtimes with verifiable contracts, evidence-gated promotions, and feedback-driven system upgrades, proven via a multi-SKU order fulfillment case study.

AI-Native DeliveryDelivery HarnessMulti-SKU Order

0 likes · 25 min read

AI-Native Delivery Evolution: Harness Framework for Controllable, Auditable Engineering

DeWu Technology

Aug 26, 2026 · Artificial Intelligence

Enterprise MultiAgent Memory: Short-Term Context & Four-Layer Architecture

The article details a four-layer memory system for enterprise MultiAgent platforms, covering short-term memory loading/writing with Redis/MySQL, long-term memory retrieval via MemOS semantic search, token budget allocation, LLM-based memory judgment, deduplication, conflict resolution, and write-before-delete strategies.

Conflict ResolutionLLM JudgmentMemOS

0 likes · 31 min read

Enterprise MultiAgent Memory: Short-Term Context & Four-Layer Architecture

DeWu Technology

Aug 19, 2026 · Artificial Intelligence

How EP-Harness Turns Personal AI Coding into a Team‑Level Agent Workflow

The article analyzes the shortcomings of using AI coding tools individually—such as unreviewed prompts, lost experience, lack of visibility, and broken development loops—and explains how EP-Harness provides a managed‑agent platform with layered architecture, unified execution contracts, context engineering, and loop automation to turn AI agents into governed, team‑wide production assets.

AI AgentsAI codingSoftware Engineering

0 likes · 12 min read

How EP-Harness Turns Personal AI Coding into a Team‑Level Agent Workflow

DeWu Technology

Aug 12, 2026 · Artificial Intelligence

Building a Judgmental RAG Agent: From Simple Search to Autonomous Search‑Filter‑Dive‑Validate Loop

This article details the design of a composite‑retrieval RAG Agent using AgentScope 2.0, covering its HarnessAgent architecture, ReAct loop, middleware‑driven three‑stage quality pipeline, multi‑source parallel search, multimodal image handling, production‑grade SSE resume, model failover, and future roadmap.

AgentScopeComposite RetrievalJava

0 likes · 23 min read

Building a Judgmental RAG Agent: From Simple Search to Autonomous Search‑Filter‑Dive‑Validate Loop

DeWu Technology

Jul 29, 2026 · Backend Development

AI‑Native Development for Transaction Core: A Five‑Gate Framework to Stabilize Legacy Systems

To adapt legacy order‑system development to AI‑generated code, the authors propose a five‑gate, spec‑driven workflow—demand clarification, technical design, TDD implementation, gate‑controlled review, and end‑to‑end telemetry—built on a five‑layer bottom‑up architecture that enforces stability, observability, and human‑validated safeguards.

AI codingBackend DevelopmentObservability

0 likes · 17 min read

AI‑Native Development for Transaction Core: A Five‑Gate Framework to Stabilize Legacy Systems

DeWu Technology

Jul 23, 2026 · Artificial Intelligence

When Engineers Cross Boundaries: How “Boundary‑Breaking” Boosted Problem Solving at Dewu

The Dewu tech team’s recent “boundary‑crossing” incidents—engineers skipping formal specs to talk directly with users and operations—led to deeper problem understanding, rapid prototyping, and measurable improvements such as 60% automated answers, 30% support load reduction, and 50% faster onboarding.

AI assistantAI platformLLM

0 likes · 7 min read

When Engineers Cross Boundaries: How “Boundary‑Breaking” Boosted Problem Solving at Dewu

DeWu Technology

Jul 22, 2026 · Artificial Intelligence

Understanding RAG: Chunking, Embedding, Similarity, HNSW and Multi‑Path Retrieval

This article explains why Retrieval‑Augmented Generation is needed, walks through the offline and online pipeline, details knowledge sources, embedding models, chunking strategies, similarity metrics, the HNSW index, multi‑path recall, RRF fusion and cross‑encoder reranking, and summarizes practical takeaways for building effective RAG systems.

ChunkingCross-EncoderEmbedding

0 likes · 36 min read

Understanding RAG: Chunking, Embedding, Similarity, HNSW and Multi‑Path Retrieval

DeWu Technology

Jul 13, 2026 · Artificial Intelligence

From Manual API Calls to Thinking Agents: DeWu Recommendation System Diagnosis

The article details DeWu's evolution from manual, experience‑driven recommendation troubleshooting to an AI‑powered diagnostic platform called “PushCheck”, describing its dual‑mode Highway/ATV architecture, the Story‑Skill framework, knowledge‑base integration, evolution pipeline, and a real‑world case study.

AI AgentDiagnostic ArchitectureKnowledge Base

0 likes · 20 min read

From Manual API Calls to Thinking Agents: DeWu Recommendation System Diagnosis