Chaowen Guo

Seeking software engineer job

Code Repository

Education

Working Experience

• handshake ai Create prompts and rubrics designed to capture real-world scenarios and requirements of various professional scenarios. Train AI to understand workplace context that define strong professional work. Systematically evaluate model responses across multiple quality dimensions. Review model-generated answers and compare them using structured criteria. Based on these evaluations, assign preference rankings that reflect which responses best satisfy the prompt's intent and demonstrate strong reasoning and communication. Create original, graduate-level STEM problems with complete step-by-step solutions, designed to induce substantive reasoning failures in state-of-the-art AI models. Improve AI accuracy across audio, visual, and text tasks.
• mercor Design Data-Centric Benchmarks for AI Models.

• Use ebay developer api to set webhook in ebay, when a new order created, use gmail api to send the order to dropshippers directly. https://gitlab.com/chaowenguo/mime/raw/main/ebay-email.mjs
• Use ebay developer api to build a html worksheet to store Order ID, Item and Price, Ship To, Earning, Return, Refund and Tracking informations. Also allow dropshippers to update tracking number directly. https://gitlab.com/chaowenguo/mime/raw/main/ebay-tracking.mjs
• Use ebay developer api to build a html worksheet to store client messages which allow the dropshippers to talk to clients directly. https://gitlab.com/chaowenguo/mime/raw/main/ebay-message.mjs

• Source code: https://gitlab.com/chaowenguo/audio/raw/main/tvb/tvb.py

Research Experience

• Developed quantum hydrodynamic trajectory simulation algorithms, with similar computational cost as a classical algorithm, to calculate chemical reactive scattering quantities.
• Implemented novel microcanonical statistical simulation called “phase space approximation”, to incorporate quantum structural effects into the quantum simulations described above.
• Programmed hybrid MPI-multithread code, and operated supercomputer clusters, to analyze performance of the above technique vs. the traditional exact quantum dynamics approach (discrete variable representation).
• Analyzed 2D and 3D data visualization to validate the accuracy of the novel approach.

Skills

Computer Techniques

• Programming Language: java, c++/c, python3, bash, javascript, fortran, torch, HTML, CSS
• Operating System: Unix/Linux, Windows, docker
• Library: MPI(Message Passing Interface with Multithread), Boost, MKL(Intel Math Kernel Library), BLAS, LAPACK, SCALAPACK
• Database: Postgresql, redis, CockroachDB
• Data Visualization: matplotlib
• DevOps: github action, gitlab ci/cd, azure pipeline
• Cloud Computing: azure, google cloud, oracle cloud, aws
• Web Auto Test: chrome devtools protocol
• Large Language Model: gemini, chatgpt

Numerical Analysis and Simulation

• Proficient in developing numerical solutions to systems of ordinary differential equations and partial differential equations
• Ability to solve problems involving large matrices, both dense and sparse, including both linear solves and eigenproblems, especially those requiring multi-node supercomputer clusters

Language

Projects

• Gen Ai to generate high resolution usa news video and upload to we media platform
• use gemini or chatgpt thinking mode to generate narration and promt for wan 2.2
• apply Qwen-Image to generate key frame and wan 2.2 vace to generate video to connect key frame and finally get a high resolution video as long as possible https://www.modelscope.cn/studios/chaowenguo/en/files
• reverse engineering edge tts and google tts to generate audio for narration
• deploy facebook/mms-tts-eng model to local gpu and synthesize english speech. https://www.modelscope.cn/studios/chaowenguo/mms/files
• pyav to generate caption and combine video and audio together and get a high resolution mp4
• rest api to upload to https://dailymotion.com/playlist/xavev4
• source code: https://huggingface.co/spaces/exact-railcar/en

• Gen Ai to generate high resolution video and upload to we media platform with much lower gpu memory usage
• use animatediff to generate video with finetune lora and prompt weight which require much lower memory than wan 2.2 and qwen-image
• deploy audioldm2 model to local gpu to generate music for video https://www.modelscope.cn/studios/chaowenguo/music/files
• patchright control chrome to upload to https://www.tiktok.com/@chaowenguo
• source code: https://www.kaggle.com/code/test6github/brabra

• web and android automation to monitor my videos
• appium to monitor my dailymotion video in android. https://dev.azure.com/chaowenguo/_git/ci?path=/browserstack.mjs
• chrome devtools protocol to control chrome in desktop to monitor my dailymotion video. https://dev.azure.com/chaowenguo/_git/ci?path=/web.mjs
• patchright control chrome in android to monitor my dailymotion video. https://dev.azure.com/chaowenguo/_git/ci?path=/android.mjs

• chrome devtools protocol to control chrome auto login freecloud and use llm rest api to bypass literal and image captcha
• https://dev.azure.com/chaowenguo/_git/ci?path=/freecloud.mjs

• deploy Qwen/Qwen3-VL-32B-Thinking llm model to local gpu with json schema output
• https://www.modelscope.cn/studios/chaowenguo/llm/files

• python Web crawler to automate web login, llm rest api to bypass literal and image captchahttps://dev.azure.com/chaowenguo/_git/ci?path=/cashmining.py
• https://bitbucket.org/chaowenguo/python-tls-client to monitor browser tls fingerprint to bypass cloudflare block