Tianyi He
@tianyihe
Senior software engineer building scalable AI, NLP, and full-stack systems for real-time impact.
What I'm looking for
I’m a Full Stack & AI Engineer with 8+ years of experience delivering scalable, production-ready software and AI systems. I’ve led end-to-end development of real-time machine learning platforms, generative AI features, and data pipelines that support billions of user interactions daily.
At Roblox, I built AI-powered content moderation pipelines using transformer-based NLP, improving moderation accuracy by 25% and enabling low-latency voice moderation with custom audio classifiers and ASR. At TikTok, I strengthened recommendations and search with GNNs, reinforcement learning, and Retrieval-Augmented Generation (RAG), driving a 15% increase in watch time and 20% higher search usage. At Tesla, I contributed to high-throughput AI data infrastructure, predictive maintenance, and CUDA-optimized inference—cutting model latency by 20% and reducing repair costs by ~15%—while consistently collaborating across product, safety, and research teams to ship responsibly.
Experience
Work history, roles, and key accomplishments
Led end-to-end development of real-time AI-powered content moderation for chat and voice, improving moderation accuracy by 25% and reducing human review dependency. Architected Kubernetes-based, GPU-accelerated inference services to achieve sub-100ms latency and 99.99% uptime during global traffic spikes, and integrated the Roblox Assistant LLM coding workflow in Roblox Studio.
Developed backend services for TikTok’s real-time content recommendation at global scale and improved the “For You” engine by integrating GNNs, increasing watch time and engagement by 15%. Built multimodal AI search with RAG over ASR-transcribed video content, raising search usage by 20%, and delivered multilingual auto-captioning across 6+ languages.
Contributed to Tesla Autopilot AI data infrastructure, processing up to 400,000 video clips/sec for neural network training and scaling distributed video ingestion and labeling workflows. Built predictive maintenance time-series anomaly detection to reduce repair costs by ~15%, optimized C++/CUDA inference to cut model latency by 20%, and improved backend microservice performance with Redis cachin
Full Stack Engineer
Supplyframe
Jun 2021 - Nov 2021 (5 months)
Built a real-time analytics dashboard for the Design-to-Source Intelligence platform using Vue.js and D3.js, enabling interactive visualization that reduced procurement decision-making time by 30%. Developed serverless AWS data processing with Lambda, DynamoDB, and API Gateway, cutting data processing latency by 40% and delivering secure, versioned REST APIs for near real-time insights.
Supported modernization of investment analysis tools by assisting in refactoring monoliths into RESTful microservices and building financial data APIs for portfolio analytics and reporting. Contributed to Morningstar Office Cloud features with responsive UI components and interactive dashboards using JavaScript, TypeScript, HTML5, and CSS3, improving usability and reliability with testing and docu
Education
Degrees, certifications, and relevant coursework
University of Southern California
Master’s Degree in Computer Science, Computer Science
2019 - 2021
Completed a Master’s degree in Computer Science at the University of Southern California from 2019 to 2021.
University of Illinois Urbana-Champaign
Bachelor’s Degree, Electrical Engineering, Statistics
2012 - 2016
Completed a Bachelor’s degree in Electrical Engineering and Statistics at the University of Illinois Urbana-Champaign from 2012 to 2016.
Tech stack
Software and tools used professionally
AWS CLI
Apache Spark
Apache Flink
D3.js
Kubernetes
MySQL
PostgreSQL
Node.js
Yarn
Nuxt.js
Unreal Engine
Redis
Vue.js
Webpack
JavaScript
HTML5
Java
Lua
MATLAB
TensorFlow
PyTorch
scikit-learn
Kafka
Amazon DynamoDB
Linux
gRPC
AWS Lambda
Serverless
Time Analytics
Assistant.to
Amazon Web Services (AWS)
CUDA
SQL
npm
Bash
Loops
100ms
Increase
Stack AI
Jan
Adjust
Availability
Location
Authorized to work in
Job categories
Skills
Interested in hiring Tianyi?
You can contact Tianyi and 90k+ other talented remote workers on Himalayas.
Message TianyiFind your dream job
Sign up now and join over 250,000+ remote workers who receive personalized job alerts, curated job matches, and more for free!
