# Ivachev Fedor (费杰)
## Executive Summary
* My current research is focused on the area of **Human-Computer Interaction** - Real world context-based conversations
* Proficient in Python, Java, C#, C++, with experience in writing code in many other languages as well
* Tools and Libraries: Git, OpenCV, NumPy, TensorFlow, pandas, Unity, IoT automation platforms, Azure API, OpenAI API, Huawei Mobile Services, Android Development
## Work Experience
#### Tsinghua University, Beijing (Teaching Assistant, 2021/01 - 2022/01)
* Assisted lead instructor in facilitating Human-Computer Interaction course
* Helped and taught both international and local students
* Developed educational tutorials and provided assistance to students for group projects
#### Huawei Technologies Co., Ltd, Moscow (Software Engineer, 2019/02 - 2019/08)
* Worked on low-level Graphic Profiling Tools using **C++**
* Retrieved information from integrated GPU counters, which helped to gain more precise timings for draw calls
* These tools were used by Graphic optimization and QA teams
* Coordinated with developers across Russia and China
#### Samsung RC, Ltd, Moscow (Intern, Programmer, 2018/06 - 2018/09)
* Developed Computer vision algorithms in 3D Avatar team using **C++** and **Python**
* Worked on OpenCV source code for removing camera distortion and implemented face physical landmark detection algorithm
* Our solution was later included into the Operating System release
## Education
* PhD in Computer Science (2021/09-present), Tsinghua University, China
* Master's in Computer Science (2019/09-2021/06), GPA 3.9, Tsinghua University, China
* Bachelor's in Computer Science (2015/09-2019/06), Lomonosov Moscow State University, Russia
## Projects + Technology stack
* **OpenAI API + Android Studio + Java + Azure + Huawei Mobile Services -** AI real-world agent - GPT-4-based agent, which utilizes multimodal information derived from Augmented Reality (AR) glasses sensors to construct comprehensive human portraits and manage everyday conversations while simultaneously providing automatic captions for unobserved data
* **Unity C# + IoT platforms -** [NUIX-Studio APP](https://github.com/FedorIvachev/NUIX-Studio-APP) - A platform to test AIoT environments in Virtual Reality. Based on Microsoft MRTK, OpenHAB and HomeAssistant (version 1), it lets users to extend real-world IoT devices with additional functionality in the Virtual Reality. Vesrion 2 is based on Oculus SDK and includes a user interface, which allows prototyping IoT interactions in Virtual Reality without any programming knowledge
* **Blender** - Special effects for TikTok - I created 10 advanced AR effects for TikTok (Douyin), which have been used for more than 3 million times.
## Qualifications and Awards
* TCL RayNeo AI Challenge - 3rd Place (2024)
* Huawei Certification: HCIA-**AI** (2020)
* MSU Higher School of Business - Best Startup - **AI** Vehicle Damage Assessment (2018)
* Certificate of 2-week Practice in **Web Apps** Development, Samsung RC, Ltd, Moscow (2017)
* Yandex.Money Hackathon - Most Technological Bot (2016)
* National and First-level Olympiads of **Mathematics** (2015, 2012), National Olympiad of **Competitive Programming** (2014)
## Publications
* Modification of Algorithm for inverting matrices with elements from ring of scalar differential operators. CMC MSU MAKS PRESS 2019: 88 (in Russian)
* ClarifAI: Context-Aware Multimodal Ontological Annotation and Abnormality Detection Through Human-AI Collaboration. IMWUT 2024 (CCF A top-level conference; under submission)
* SightTalk: Real-time Unseen Visual Data Captioning Through Human-AI Conversation. IMWUT 2024 (CCF A top-level conference; under submission)
## Misc. Information
* Languages: English (C2, IELTS 7.5), Chinese (B2, HSK3 295/300), Russian (native)
* Holder of the Hong Kong ID card, allowing right to abode without any visa restrictions.
Printed version - See the most recent version at fedorivachev.github.io