Mingjing Wu-Machine Learning Engineer | Inference Systems & Recommendation
Check rate
Experience
Software Engineer, Inference
Confidential AI Company
- Built a production inference engine from the ground up and implemented high-performance GEMM, sparse-attention, and fused kernels, improving inference performance by up to 200%.
- Enabled and optimized inference for 10+ newly released model families, including DeepSeek, Kimi, and MiniMax, improving production readiness across latency, throughput, and compute efficiency.
Machine Learning Engineer Intern
Alibaba Cloud
- Built an unsupervised NLP pipeline for 30,000 customer-service tickets using StructBERT, Siamese UniNLU, regex labeling, and Neo4j; achieved 97% accuracy and increased the automated resolution rate by 21% after deployment on Alibaba EAS.
- Developed a Prophet-based replenishment system and dynamic inventory model for cloud assemble-to-order operations; reached 98% adoption and projected RMB 100M in annualized inventory cost savings.
Search Ads Machine Learning Engineer Intern
JD Retail (JD.com)
- Built a daily BERT-to-FastText knowledge-distillation pipeline with EM-based transfer and traffic-aware pruning for query category prediction; increased impressions 1.9% and content consumption 2.2%.
- Redesigned a 6,000+ category multi-label intent classifier using Circle Loss and shared-encoder dual towers with cross/self-attention; lifted impressions 2.0% and content consumption 1.7%.
- Optimized heterogeneous-graph retrieval with MetaPath2Vec, HAN, and layer-wise training for billion-scale graphs; increased content consumption 3.1% in 1.5-2 week online A/B tests serving 30M DAU.
Recommendation Systems Engineer Intern
Kuaishou
- Enhanced LightGBM multi-objective ranking with personalized recall, PageRank, ADASYN, and Bayesian smoothing; improved offline AUC 6.9% on high-frequency traffic and 3.0% overall.
- Deployed DeepFFM/MMoE models in C++ with BPR loss and attention pooling, improving gAUC 9% relative; applied doubly robust estimation with DID and propensity-score matching to increase retention 0.7% relative.
Industry Experience
See where this freelancer has spent most of their professional time.
Experienced in Information Technology, Retail, and Media and Entertainment.
Business Area Experience
See which departments and functions this freelancer has contributed to most.
Experienced in Information Technology, Product Development, Customer Service, Logistics, Operations, and Supply Chain Management.
Summary
I build machine learning systems that improve both model quality and production efficiency, with hands-on work across inference, recommendation, search ads, and applied NLP. I built a production inference engine from the ground up and implemented high-performance GEMM, sparse-attention, and fused kernels, improving inference performance by up to 200%, while enabling 10+ newly released model families for production.
I also develop practical ML solutions that move business metrics. My work has improved latency, throughput, impressions, content consumption, retention, and automated resolution rates through recommendation models, graph retrieval, knowledge distillation, and NLP pipelines. My background in analytics and mathematics helps me turn complex models into reliable systems that perform well at scale.
Skills
- Programming: Python, C++, Java
- Ml Systems: Distributed Inference, Kernel Optimization, Model Serving, Profiling/Benchmarking
- Ml: Retrieval, Ranking, Personalization, Knowledge Distillation, A/B Testing
Languages
Education
Nanyang Technological University
M.S. · Analytics · Singapore, Singapore · 4.3/5.0
Central South University
B.S. · Mathematics and Applied Mathematics · Chang Sha Shi, China · 3.7/4.0
Statistics
Experience
Expertise
Qualifications
Profile
Frequently asked questions
Have questions? Find more information here.
Daily Rate Distribution
The rates shown represent the typical market range for freelancers in this position based on recent contracts on our platform.
Actual rates may vary depending on seniority level, experience, skill specialization, project complexity, and engagement length.
Average rates for similar positions
Rates are based on recent contracts and do not include FRATCH margin.
The average daily rate is the mean of all daily rates from recent contracts of comparable freelancers on our platform.
The median daily rate is the middle value of all daily rates — half of comparable freelancers charge less, half charge more. Unlike the average, it is barely affected by outliers.
Actual rates may vary depending on seniority level, experience, skill specialization, project complexity, and engagement length.
Similar Freelancers
Discover other experts with similar qualifications and experience
FRATCH Freelancer
David Dai
Qin Zheng
Julia Lach
Borui Freelancer
Tianhui Zou
Jonathan Staude
Keying Freelancer
Marcel Steger
Tobias Jaeuthe
Experts recently working on similar projects
Freelancers with hands-on experience in comparable project as a Software Engineer, Inference
Matthias Isler
Ursula Maria Mayer
Manuel Pasieka
David Lynn
Kai Wolf
Lucas Habrich
Lino Giefer
Florin Dorel
Shiqing Fan
Stephan Fröde
Nearby freelancers
Professionals working in or nearby Singapore, Singapore
Dinmukhabet Anuarbekov
Roman Sayapin
Sergey Filin
Hikmat Shirmammadov
Javid Karimov
Emin Maharramli
Beka Chogovadze
Revaz Todadze
Konstantin Freelancer
Lizi Shavadze
Most recent projects
FRATCH works with many companies and recruitment agencies. Here you will find our recently posted projects and opportunities.
