
Can Li
Senior Software Engineer @ Meta
About
Senior Software Engineer at Meta with over five years of experience specializing in AI/ML Infrastructure and Recommendation Systems. I focus on advancing recommender technologies for flagship products such as Instagram Reels and Facebook Feed. I played a key role in designing critical components of the FB Feed Recommendation, driving measurable improvements in user engagement, capacity optimization, and enabling the successful launch of 30+ features. I have led multiple high-impact AI/ML infrastructure initiatives, including the development of a scalable service for generating/validating sequence-based training data for Ads ranking foundation models—resulting in global model performance gains and significant resource efficiencies. Additionally, I spearheaded projects to enhance topical model accuracy by implementing innovative strategies to improve training data quality. Recognized for strong cross-functional collaboration, strategic problem-solving, and a commitment to aligning technical solutions with business objectives. My deep domain expertise helps power Meta’s mission to build meaningful connections and ensure safer, more engaging user experiences at scale.
United States
New York City Metropolitan Area
Computer Software
Machine Learning, Artificial Intelligence (AI), Recommender Systems, C++, Python, C, Java, Programming, OpenGL, CUDA, Algorithms, Computer Science, Qt, Real-time Control Systems, Embedded C, Embedded Systems, Software Development, Simulation Software, MFC, Software Engineering
Experience

Senior Software Engineer
New York City Metropolitan Area
Led multiple Recommendation System projects, supporting flagship products (IG Reels, FB Feed Recs) - Designed and developed a critical component in FB Feed Recommendation, driving significant topline gains (+0.0X% App Sessions, +0.0X% Feed Views, capacity savings) and enabling 30+ feature launches over 1.5 years. - Designed and built integrity measurement components for IG Reels, delivering key metrics on violating content prevalence and guardrails to support safe model launches Led multiple AI/ML Infra projects to support ranking models powering RecSys and Ads - Designed and developed a service to generate/validate training data for Ads ranking foundation models, leading to +%0.X Global Ads Score and $1.XM capacity savings. - Designed and developed a heuristic-based approach to evaluate human labeler performance and detect bad labels in training data, resulting in improved R@P70 of multiple topical models. - Designed and automated e2e labeling workflows to evaluate topical model performance and assess human labeler quality at scale

Software Engineer
Greater San Diego Area
EMBEDDED SYSTEM | SOFTWARE DEVELOPMENT ♦ Maintained& Developed real-time control software with C/C++, Linux, Python ♦ Developed a hardware simulation software based on B&R PLC with C/C++ through analyzing hardware specification, modeling communication protocol and simplifying core functionality ♦ Bug-fixing and performance-improvement on legacy code with Agile project management ♦ Coordinated & performed software test with different domain experts in the field ♦ Technical documents reiview, writing and presentation

Graduate Teaching Assistant
Houghton, Michigan
COMPUTER PROGRAMMING | TEACHING & DEVELOPMENT ♦ Responsible for providing homework assignment support and examination preparations while leveraging programming knowledge to support education regarding Introduction to Computing (100+ students) and System Programming (60+ students) ♦ Assist with implementation of computer programming lesson plans that met varied student needs, evaluated and assessed students’ development, and implemented study methodologies to enhance students’ progress

★ Research Assistant
Houghton, Michigan
PROJECT MANAGEMENT | RESEARCH AND ANALYSIS ♦ Demonstrated proficiency and efficiently utilized skills in C++, Qt, and Linux, to successfully accomplish Visualization Software Project for Vigenére Cipher ♦ Implemented demo mode to display the encryption and decryption process in (3) tools, and implmeneted the practice mode to allow users to practice doing the encryption and decryption ♦ Implemented Kasiski test & IoC method to analyze the keyword length, and implemented X2 method to recover the keyword ♦ Expressed a full understand of project requirements in alignment with targeted software development

★ Research Assistant
Michigan Technological University
Houghton, Michigan
PROJECT MANAGEMENT | RESEARCH AND ANALYSIS ♦ Leveraged skills in C++, CUDA, OpenGL, and Linux technologies to visualize unsteady flow fields with pathline and accomplish 4 critical steps (entropy computation, critical region detection, skeleton construction, and path-line tracing) regarding the Visualization of Unsteady Flow Field project ♦ Implemented the entropy computation with CUDA to extract critical regions based on Out-of-Core technique ♦ Implemented Volume Thinning Algorithm to generate skeletons of the critical regions ♦ Designed and implemented an algorithm to construct spatial-temporal regions ♦ Implemented the pathline-tracing algorithm by placing seeds on the skeletons of the flow field

★ Research Assistant
China
PROJECT MANAGEMENT | RESEARCH & ANALYSIS ♦ Demonstrated proficiency with C++, CUDA, and OpenGL for Volume Rendering of Electromagnetic Field project ♦ Visualized electromagnetic information through the volume-rendering method, and devised differed algorithms ♦ Designed and implemented parallel algorithm to calculate the electromagnetic value of each discrete point in the 3D world with CUDA ♦ Implemented the ray-casting parallel algorithm with CUDA to visualize data and render the electromagnetic field ♦ Implemented Empty-Space Skipping Algorithm to improve the efficiency of rendering based on Octree ♦ Designed and implemented the Matrix Generation Algorithm to merge the ray-casting result into the 3D scene
Education

Applied Physics
Selected Courses: Calculus I&II, Linear Algebra, Probability and Mathematical Statistics, Data Structure, C Programming, Numerical Computing Methods, Principle of Microcomputer System and Interface, Matlab Experiment
Can Li's Contact Information
Phone
Find the Right Leads
Find Verified Contact Data
What LeadContact does well
Find verified emails, phone numbers, and decision-makers with 98% accuracy.
Find Leads
Find the right people by company, role, industry, location, and more.
925M+ professional profiles

Find Emails
Access verified email addresses for your target contacts.
657M+ emails

Find Phone Numbers
Get cross-validated phone data from multiple top sources.
239M+ phone numbers

More Accurate. Lower Cost.
Find contact data in 1 tool with 98% accuracy
LeadContact integrates leading enrichment tools to deliver more accurate contact data—without paying for each one.
Great conversations start with the right contact.
It’s time to find yours.




