diff --git a/README.md b/README.md index 9d45fa4..0d61f83 100644 --- a/README.md +++ b/README.md @@ -5,202 +5,56 @@ - Website: [miti99.com](https://miti99.com/) - LinkedIn: [miti99](https://linkedin.com/in/miti99) - GitHub: [tiennm99](https://github.com/tiennm99) +- Instagram: [tiennm99](https://instagram.com/tiennm99) -# Welcome to RenderCV -RenderCV reads a CV written in a YAML file, and generates a PDF with professional typography. - -See the [documentation](https://docs.rendercv.com) for more details. - # Education -## **Princeton University**, PhD in Computer Science -- Princeton, NJSept 2018 – May 2023 - -- Thesis: Efficient Neural Architecture Search for Resource-Constrained Deployment - -- Advisor: Prof. Sanjeev Arora - -- NSF Graduate Research Fellowship, Siebel Scholar (Class of 2022) - - - -## **Boğaziçi University**, BS in Computer Engineering -- Istanbul, TürkiyeSept 2014 – June 2018 - -- GPA: 3.97/4.00, Valedictorian - -- Fulbright Scholarship recipient for graduate studies +## **Ho Chi Minh City University of Technology**, B.E. in Computer Science in Computer Science and Engineering -- Ho Chi Minh City, VietnamSept 2017 – June 2023 # Experience -## **Co-Founder & CTO**, Nexus AI -- San Francisco, CA +## **Senior Software Engineer**, ZingPlay Game Studios, VNG Corporation -- Ho Chi Minh City, Vietnam -June 2023 – present +July 2020 – present -- Built foundation model infrastructure serving 2M+ monthly API requests with 99.97% uptime +Started my journey at VNG Tech Fresher Program and progressed to Senior Software Engineer at ZingPlay Game Studios (ZPS). Over the years, I have honed my expertise in game server architecture and backend development using Java, while also contributing to client-side logic with Cocos and Godot when needed. -- Raised $18M Series A led by Sequoia Capital, with participation from a16z and Founders Fund +- [Show](https://play.google.com/store/apps/details?id=zps.games.show) -- Scaled engineering team from 3 to 28 across ML research, platform, and applied AI divisions + - A card game for Myanmar market -- Developed proprietary inference optimization reducing latency by 73% compared to baseline +- [Burkozel](https://play.google.com/store/apps/details?id=zps.games.burkozel) + - A card game for the Russian audience +- [Bida3D](https://play.google.com/store/apps/details?id=zps.games.bida3d.vn) -## **Research Intern**, NVIDIA Research -- Santa Clara, CA + - Global 8-ball pool game -May 2022 – Aug 2022 +- [Chaos Age 2](https://play.google.com/store/apps/details?id=vn.zps.tl2) -- Designed sparse attention mechanism reducing transformer memory footprint by 4.2x - -- Co-authored paper accepted at NeurIPS 2022 (spotlight presentation, top 5% of submissions) - - - -## **Research Intern**, Google DeepMind -- London, UK - -May 2021 – Aug 2021 - -- Developed reinforcement learning algorithms for multi-agent coordination - -- Published research at top-tier venues with significant academic impact - - - ICML 2022 main conference paper, cited 340+ times within two years - - - NeurIPS 2022 workshop paper on emergent communication protocols - - - Invited journal extension in JMLR (2023) - - - -## **Research Intern**, Apple ML Research -- Cupertino, CA - -May 2020 – Aug 2020 - -- Created on-device neural network compression pipeline deployed across 50M+ devices - -- Filed 2 patents on efficient model quantization techniques for edge inference - - - -## **Research Intern**, Microsoft Research -- Redmond, WA - -May 2019 – Aug 2019 - -- Implemented novel self-supervised learning framework for low-resource language modeling - -- Research integrated into Azure Cognitive Services, reducing training data requirements by 60% + - Global strategy game # Projects -## **[FlashInfer](https://github.com/)** +## **[Static websites with Hugo](https://tiennm99.github.io/)** -Jan 2023 – present +Jan 2020 – present -Open-source library for high-performance LLM inference kernels - -- Achieved 2.8x speedup over baseline attention implementations on A100 GPUs - -- Adopted by 3 major AI labs, 8,500+ GitHub stars, 200+ contributors +My blog on GitHub Pages using Hugo. Website for Ngăm - a charity project founded by my brother's friends. -## **[NeuralPrune](https://github.com/)** - -Jan 2021 - -Automated neural network pruning toolkit with differentiable masks - -- Reduced model size by 90% with less than 1% accuracy degradation on ImageNet - -- Featured in PyTorch ecosystem tools, 4,200+ GitHub stars - - - -# Publications -## **Sparse Mixture-of-Experts at Scale: Efficient Routing for Trillion-Parameter Models** - -July 2023 - -*John Doe*, Sarah Williams, David Park - -[10.1234/neurips.2023.1234](https://doi.org/10.1234/neurips.2023.1234) (NeurIPS 2023) - - - -## **Neural Architecture Search via Differentiable Pruning** - -Dec 2022 - -James Liu, *John Doe* - -[10.1234/neurips.2022.5678](https://doi.org/10.1234/neurips.2022.5678) (NeurIPS 2022, Spotlight) - - - -## **Multi-Agent Reinforcement Learning with Emergent Communication** - -July 2022 - -Maria Garcia, *John Doe*, Tom Anderson - -[10.1234/icml.2022.9012](https://doi.org/10.1234/icml.2022.9012) (ICML 2022) - - - -## **On-Device Model Compression via Learned Quantization** - -May 2021 - -*John Doe*, Kevin Wu - -[10.1234/iclr.2021.3456](https://doi.org/10.1234/iclr.2021.3456) (ICLR 2021, Best Paper Award) - - - -# Selected Honors -- MIT Technology Review 35 Under 35 Innovators (2024) - -- Forbes 30 Under 30 in Enterprise Technology (2024) - -- ACM Doctoral Dissertation Award Honorable Mention (2023) - -- Google PhD Fellowship in Machine Learning (2020 – 2023) - -- Fulbright Scholarship for Graduate Studies (2018) - # Skills -**Languages:** Python, C++, CUDA, Rust, Julia +**Programming:** Java (Netty, Vert.x, Spring Boot), JavaScript -**ML Frameworks:** PyTorch, JAX, TensorFlow, Triton, ONNX +**Databases:** Couchbase, Redis, MySQL -**Infrastructure:** Kubernetes, Ray, distributed training, AWS, GCP +**Tools & DevOps:** Git, Docker, CI/CD -**Research Areas:** Neural architecture search, model compression, efficient inference, multi-agent RL +# Interests +**Professional:** Game server architecture, distributed systems, Java performance tuning -# Patents -1. Adaptive Quantization for Neural Network Inference on Edge Devices (US Patent 11,234,567) - -1. Dynamic Sparsity Patterns for Efficient Transformer Attention (US Patent 11,345,678) - -1. Hardware-Aware Neural Architecture Search Method (US Patent 11,456,789) - -# Invited Talks -1. Scaling Laws for Efficient Inference — Stanford HAI Symposium (2024) - -1. Building AI Infrastructure for the Next Decade — TechCrunch Disrupt (2024) - -1. From Research to Production: Lessons in ML Systems — NeurIPS Workshop (2023) - -1. Efficient Deep Learning: A Practitioner's Perspective — Google Tech Talk (2022) - -# Any Section Title -You can use any section title you want. - -You can choose any entry type for the section: `TextEntry`, `ExperienceEntry`, `EducationEntry`, `PublicationEntry`, `BulletEntry`, `NumberedEntry`, or `ReversedNumberedEntry`. - -Markdown syntax is supported everywhere. - -The `design` field in YAML gives you control over almost any aspect of your CV design. - -See the [documentation](https://docs.rendercv.com) for more details. +**Personal:** Reading novels & manga, playing Genshin Impact & TFT diff --git a/index.html b/index.html index 2380819..45b6cbe 100644 --- a/index.html +++ b/index.html @@ -46,194 +46,51 @@
RenderCV reads a CV written in a YAML file, and generates a PDF with professional typography.
-See the documentation for more details.
Thesis: Efficient Neural Architecture Search for Resource-Constrained Deployment
-Advisor: Prof. Sanjeev Arora
-NSF Graduate Research Fellowship, Siebel Scholar (Class of 2022)
-GPA: 3.97/4.00, Valedictorian
-Fulbright Scholarship recipient for graduate studies
-June 2023 – present
+July 2020 – present
+Started my journey at VNG Tech Fresher Program and progressed to Senior Software Engineer at ZingPlay Game Studios (ZPS). Over the years, I have honed my expertise in game server architecture and backend development using Java, while also contributing to client-side logic with Cocos and Godot when needed.
Built foundation model infrastructure serving 2M+ monthly API requests with 99.97% uptime
+Raised $18M Series A led by Sequoia Capital, with participation from a16z and Founders Fund
+A card game for Myanmar market
Scaled engineering team from 3 to 28 across ML research, platform, and applied AI divisions
+Developed proprietary inference optimization reducing latency by 73% compared to baseline
-May 2022 – Aug 2022
-Designed sparse attention mechanism reducing transformer memory footprint by 4.2x
+A card game for the Russian audience
Co-authored paper accepted at NeurIPS 2022 (spotlight presentation, top 5% of submissions)
-May 2021 – Aug 2021
-Developed reinforcement learning algorithms for multi-agent coordination
+Published research at top-tier venues with significant academic impact
+Global 8-ball pool game
ICML 2022 main conference paper, cited 340+ times within two years
+NeurIPS 2022 workshop paper on emergent communication protocols
-Invited journal extension in JMLR (2023)
-May 2020 – Aug 2020
-Created on-device neural network compression pipeline deployed across 50M+ devices
-Filed 2 patents on efficient model quantization techniques for edge inference
-May 2019 – Aug 2019
-Implemented novel self-supervised learning framework for low-resource language modeling
-Research integrated into Azure Cognitive Services, reducing training data requirements by 60%
+Global strategy game
Jan 2023 – present
-Open-source library for high-performance LLM inference kernels
-Achieved 2.8x speedup over baseline attention implementations on A100 GPUs
-Adopted by 3 major AI labs, 8,500+ GitHub stars, 200+ contributors
-Jan 2021
-Automated neural network pruning toolkit with differentiable masks
-Reduced model size by 90% with less than 1% accuracy degradation on ImageNet
-Featured in PyTorch ecosystem tools, 4,200+ GitHub stars
-July 2023
-John Doe, Sarah Williams, David Park
-10.1234/neurips.2023.1234 (NeurIPS 2023)
-Dec 2022
-James Liu, John Doe
-10.1234/neurips.2022.5678 (NeurIPS 2022, Spotlight)
-July 2022
-Maria Garcia, John Doe, Tom Anderson
-10.1234/icml.2022.9012 (ICML 2022)
-May 2021
-John Doe, Kevin Wu
-10.1234/iclr.2021.3456 (ICLR 2021, Best Paper Award)
-MIT Technology Review 35 Under 35 Innovators (2024)
-Forbes 30 Under 30 in Enterprise Technology (2024)
-ACM Doctoral Dissertation Award Honorable Mention (2023)
-Google PhD Fellowship in Machine Learning (2020 – 2023)
-Fulbright Scholarship for Graduate Studies (2018)
-Jan 2020 – present
+My blog on GitHub Pages using Hugo. Website for Ngăm - a charity project founded by my brother's friends.
Languages: Python, C++, CUDA, Rust, Julia
-ML Frameworks: PyTorch, JAX, TensorFlow, Triton, ONNX
-Infrastructure: Kubernetes, Ray, distributed training, AWS, GCP
-Research Areas: Neural architecture search, model compression, efficient inference, multi-agent RL
-Adaptive Quantization for Neural Network Inference on Edge Devices (US Patent 11,234,567)
-Dynamic Sparsity Patterns for Efficient Transformer Attention (US Patent 11,345,678)
-Hardware-Aware Neural Architecture Search Method (US Patent 11,456,789)
-Scaling Laws for Efficient Inference — Stanford HAI Symposium (2024)
-Building AI Infrastructure for the Next Decade — TechCrunch Disrupt (2024)
-From Research to Production: Lessons in ML Systems — NeurIPS Workshop (2023)
-Efficient Deep Learning: A Practitioner's Perspective — Google Tech Talk (2022)
-You can use any section title you want.
-You can choose any entry type for the section: TextEntry, ExperienceEntry, EducationEntry, PublicationEntry, BulletEntry, NumberedEntry, or ReversedNumberedEntry.
Markdown syntax is supported everywhere.
-The design field in YAML gives you control over almost any aspect of your CV design.
See the documentation for more details.
+Programming: Java (Netty, Vert.x, Spring Boot), JavaScript
+Databases: Couchbase, Redis, MySQL
+Tools & DevOps: Git, Docker, CI/CD
+Professional: Game server architecture, distributed systems, Java performance tuning
+Personal: Reading novels & manga, playing Genshin Impact & TFT