Deng Yaqi

AI R&D Lead / RAG and Agent Engineering / Backend Systems

I am Deng Yaqi. I currently lead AI R&D planning, technical direction, and delivery. My path started with NLP, knowledge graphs, and ontology fusion, then moved through Android, Java backend systems, risk-control platforms, and distributed systems, before returning to enterprise AI platforms, RAG, Agent workflows, and production delivery.

Adventure Notes

Outside of work, I like going farther into mountains, lakes, and the edges of cities, then bringing that sense of exploration back into engineering judgment.

Snow mountains, a lake, and a traveler seen from behind

Work Experience

AI R&D Lead

Hangzhou Genaima Network Technology Co., Ltd. / Hangzhou, China

2026.04 - Present

Leads AI R&D planning, technical direction, and project delivery, bringing enterprise AI platforms, Agent/RAG workflows, data governance, and business copilots from zero to one.

Senior Software Development Engineer

iSoftStone Technology Service Shanghai Co., Ltd. / Shanghai, China

2025.05 - 2026.04

Developed RAG systems, vector knowledge bases, intelligent Q&A engines, and AI coding platforms, moving AI tools from proofs of concept toward deliverable applications.

Senior Java Developer

Beijing Hanke Times Technology Co., Ltd. / Ant Group MYbank Risk Control

2023.06 - 2024.10

Worked onsite with Ant Group's MYbank risk-control team, developing backend capabilities for strategy R&D, industry risk management, and risk strategy platforms.

Android Software Engineer

Zhejiang Tsinghua Flexible Electronics Technology Institute / Hangzhou, China

2021.11 - 2022.12

Built medical Android software end to end, covering design, coding, debugging, device interaction, cloud communication, and post-release support.

Freelancer

Paris, France

2019.10 - 2021.08

Delivered project-based technical solutions across NLP algorithms, Java web applications, and Android software.

Natural Language Processing Engineer

Yseop / Paris, France

2019.04 - 2019.09

Researched knowledge graph projects, including ontology-fusion algorithm development and validation, and completed full-stack development for the Lexicon website.

Ask Me About

AI / RAG / Agent Engineering

LangChain LangGraph Agent Workflow MCP RAG Embedding Prompt Engineering Multi-model LLM

Backend and Distributed Systems

Java Python FastAPI Django REST Framework SpringBoot SpringCloud SofaBoot RESTful API

Data, Search, and Knowledge Graphs

PostgreSQL PgVector ClickHouse Neo4j Redis Elasticsearch MySQL HBase

Knowledge Graphs and Ontology

Knowledge Graph Ontology Ontology Fusion Information Extraction NER Data Annotation

Frontend and Browser Extensions

Vue 3 TypeScript Vite Element Plus Pinia Chrome Extension MV3 SSE

Mobile and Client Apps

Android Kotlin Jetpack MVVM Bluetooth IoT Electron

DevOps and Infrastructure

Docker Kubernetes Nginx Jenkins GitLab CI GitHub Actions PyInstaller

Tools and Collaboration

VSCode Cursor Codex Claude Code Power BI Postman Figma

Credentials and Languages

Certified

AI Trainer Level 3

Professional skill certificate

Certified

AI Trainer Level 4

Professional skill certificate

Business French foundation

DELF B2

French language certificate

Certified

C2 Driving License

China motor vehicle license

Native

Chinese

Mandarin / Cantonese / Hakka

Business / Fluent

Foreign Languages

French / English / Japanese

Life and Interests

Travel and Mountains

I like walking into mountains, lakes, and city edges, then bringing those observations back into engineering judgment.

Languages and Culture

French, English, Japanese, and cross-cultural work keep me attentive to language and knowledge expression.

Writing and Notes

I turn project reviews, technical judgment, and AI tool practice into reusable public notes.

Product Exploration

I care about how AI capabilities move from experiments into systems people can actually use.

Current Direction

Building: Enterprise RAG, document parsing quality, PgVector knowledge bases, multi-model LLM integrations, Agent workflows, and AI coding systems.

Learning: Context engineering, evaluation systems, tracing, cost control, and the reliability boundary of model capabilities in production.

Writing: Turning real engineering judgment into clear, verifiable, reusable public notes.