Peimeng Sui
Data AnalystMachine Learning @ Pinterest
Pinterest
AI is analyzing your overall score…
Identifying your key strengths…
Evaluating your skill match against the job requirements…
Assessing your cultural and operational fit
Machine Learning @ Pinterest
AI is analyzing your overall score…
Identifying your key strengths…
Evaluating your skill match against the job requirements…
Assessing your cultural and operational fit
I am a machine learning generalist passionate about AI applications for positive impact. I have experience building machine learning solutions in recommendation systems, search relevance, and computer vision. I have shipped multiple large-scale and highly impactful products, e.g., personalized recommenders at TripAdvisor, search Typeahead ranking at LinkedIn, and real-time depth estimation at Meta.
UCLA
Bachelor of Science - BS, Applied Mathematics & Statistics
N/A – Present
New York University
Master of Science - MS, Data Science
N/A – Present
Staff Machine Learning Engineer
August 1, 2024 – Present
Meta
Software Engineer, Machine Learning
September 1, 2022 – August 1, 2024
Senior Software Engineer - Data Mining/Data Analysis/Machine Learning
April 1, 2020 – September 1, 2022
Tripadvisor
Senior Machine Learning Scientist
March 1, 2020 – April 1, 2020
Greater Boston Area
Tripadvisor
Machine Learning Scientist II
March 1, 2019 – March 1, 2020
Greater Boston Area
Tripadvisor
Machine Learning Scientist
February 1, 2018 – March 1, 2019
Greater Boston Area
Imagen Technologies
Machine Learning Research Intern
September 1, 2017 – December 1, 2017
Greater New York City Area
TripAdvisor
Data Sceince Intern
June 1, 2017 – August 1, 2017
Greater Boston Area
Space Jam Data
Data Science Intern
September 1, 2016 – December 1, 2016
Greater New York City Area
The Chinese University of Hong Kong
Summer Research Intern at Computer Science Engineering Department
June 1, 2015 – August 1, 2015
Hong Kong
VCLA (Center of Vision, Cognition, Learning and Art)
Undergraduate Researcher
February 1, 2015 – March 1, 2016
Semi-supervised MNIST digit recognition
February 1, 2017 – March 1, 2017
This is the group project for in-class Kaggle competition MNIST_DIGIT_RECOGNITION_ASSIGNMENT_1. The class is DS-GA 1008 Deep Learning at NYU, provided by Yann Lecun. The problem setting for this competition is Semi-Supervise Learning. We have 3,000 labeled samples and 47,000 unlabeled samples. Using ConvNets, along with data augmentation, pseudo-label method, and ensembling, we get a 99.42% accuracy on 10,000 test samples, ranking the 2nd out of 30 teams.
Computer Vision Course Project: Neural style transfer
November 1, 2016 – Present
Implementing and enhancing paper “A Neural Algorithm of Artistic Style” in Torch
Computer Vision Kaggle Challenge Traffic sign competition
November 1, 2016 – Present
Image Classification using Convolutional Neural Network implemented in Torch Achieve the highest test set accuracy on the German Traffic Sign Recognition Benchmark
Intro to Data Science Course Project: Allstate Claims Severity Kaggle Challenge
November 1, 2016 – Present
Building a pipeline of problem formulation, data prep/understanding, modeling and discussion on implementation.
Kaggle Project Classification of Animal Outcomes Using Tree-based Methods
May 1, 2016 – Present
1. Applied Rnadom Forest and Boosting on 113,891 records of training dataset 2. Created new features (17 old variables to 31 new variables) and explored external data 3. Achieved overall prediction error rate 12% (1st place on kaggle final leaderboard among 55 teams)
Detection of Newsgroup Label Using SVM
May 1, 2016 – Present
In this project we are trying to apply machine learning technique - SVM on the 20 Newsgroup dataset, which is a collection of 18774 newsgroup documents. The data is organized into 20 different topics. As a conclusion of this project, we find that 2-class SVM has over 95% testing accuracy rate without dimension reduction. Dimension Reduction by selecting important words improves efficiency a lot without losing too much prediction accuracy. Multiclass SVM works still good with over 80% prediction accuracy.
2015 USC Data Analytics Competititon
November 1, 2015 – Present
1. 5 hours intensive teamwork analyzing data provided by Paramount Pictures 2. Selected the most significant digital performance indicator using the adaptive boosting algorithm
2015 Summer Undergraduate Research Program at CUHK
June 1, 2015 – Present
Focus on development of distributed computing systems for processing big data(Supervisor: Prof. James Cheng). Learn how to use Hadoop to process big dataset and write java code in mapreduce format. Learn and implement popular machine learning algorithms(Linear Regression, Logistic Regression, KMeans, Stochastic Gradient Descent, K Nearest Neighbors, Collaborative Filter, DBSCAN, etc).
2015 ASA Datafest
April 1, 2015 – Present
Based on analysis of data provided by Edmunds.com, we came up with recommendations for the company to make the shopping process much easier.
VCLA (Center of Vision, Cognition, Learning and Art) Undergraduate Researcher
March 1, 2015 – Present
The Center for Vision, Cognition, Learning, and Art (VCLA) is affiliated with the Departments of Statistics and Computer Science at UCLA. We start from Computer Vision and expand to other disciplines. Our objective is to pursue a unified framework for representation, learning, inference and reasoning, and to build intelligent computer systems for real world applications.
Cornerstone Case Competition, Finalist
January 1, 2015 – Present
Our team has been hired by Intertrode to assess the damages for the lawsuit. We prepare a presentation detailing how much damages should be owed to Intertrode and how you arrived at your conclusion.
Deloitte Case Competition, Finalist
January 1, 2015 – February 1, 2015
We come up with the optimum solution involving strategic, technological and human capital innovations to adapt to the changed market in glasses retail industry.
Cultural Fit Analysis
The candidate's extensive experience across multiple prominent tech companies (Meta, LinkedIn, Pinterest, Tripadvisor) and participation in various academic and industry competitions demonstrates adaptability and a proactive learning attitude. The diverse range of projects, from computer vision to recommendation systems and natural language processing, indicates a broad interest in data science and machine learning domains. While the target role is 'Data Analyst', the candidate's background is heavily skewed towards Machine Learning Engineering and Research, which might indicate a potential mismatch in day-to-day responsibilities if the Data Analyst role is purely focused on reporting and descriptive analytics. However, their strong analytical and data processing skills are highly relevant.
Soft Skills & Operational Fit
The candidate's project descriptions and work experience highlight strong problem-solving, analytical thinking, and a results-oriented approach. Their involvement in team-based competitions and leading initiatives suggests good collaboration and leadership potential. The focus on deploying models and observing business impact indicates a practical, operational mindset.