cv.tex

%-------------------------
% Resume in Latex
% Author : Nikan Doosti
%------------------------

\documentclass[letterpaper,11pt]{article}

% \include{command.tex}

\usepackage{latexsym}
\usepackage[empty]{fullpage}
\usepackage{titlesec}
\usepackage{marvosym}
\usepackage[usenames,dvipsnames]{color}
\usepackage{verbatim}
\usepackage{enumitem}
\usepackage{hyperref}
\usepackage{fancyhdr}
\usepackage{array}
\usepackage{graphicx}
\usepackage{wrapfig}


\pagestyle{fancy}
% \fancyfoot{}
\fancyhf{}  % Clear default settings
\fancyfoot[R]{\small \emph{Nikan Doosti} \hspace{1em} \thepage}
\setlength{\footskip}{26pt}  % Adjusts the distance between text and footer
\fancyhfoffset[R]{0.5cm}  % Pushes the footer content further to the right

\renewcommand{\headrulewidth}{0pt}
\renewcommand{\footrulewidth}{0pt}

% Adjust margins
\addtolength{\oddsidemargin}{-0.375in}
\addtolength{\evensidemargin}{-0.375in}
\addtolength{\textwidth}{1in}
\addtolength{\topmargin}{-.5in}
\addtolength{\textheight}{1.0in}

\urlstyle{same}

\raggedbottom
\raggedright
\setlength{\tabcolsep}{0in}

% Sections formatting
\titleformat{\section}{
  \vspace{-4pt}\scshape\raggedright\large
}{}{0em}{}[\color{black}\titlerule \vspace{-5pt}]

%-------------------------
% Custom commands
\newcommand{\resumeItem}[1]{
  \item\small{
    {#1 \vspace{0pt}}
  }
}

\newcommand{\LinedItem}[1]{\item \begin{tabular}[t]{|p{10cm}}#1\end{tabular}}

\newcommand{\resumeSubheading}[5]{
  \item
    \begin{tabular*}{0.97\textwidth}{l@{\extracolsep{\fill}}r}
      \textbf{#1} & #2 \\ 
      \textit{\small#3} & \textit{\small #4} \\
      \textit{#5} & 
    \end{tabular*}\vspace{-5pt}
}

\newcommand{\resumeSubheadingF}[4]{
  \item
    \begin{tabular*}{0.97\textwidth}{l@{\extracolsep{\fill}}r}
      \textbf{#1} & #2 \\ 
      \textit{\small#3} & \textit{\small #4} \\
    \end{tabular*}\vspace{-5pt}
}

\newcommand{\awarditem}[2]{
  \item
    \begin{tabular*}{0.97\textwidth}{l@{\extracolsep{\fill}}r}
      #1 & #2 \\ 
    \end{tabular*}\vspace{-5pt}
}


\newcommand{\resumeSubItem}[2]{\resumeItem{#1}{#2}\vspace{-4pt}}

\renewcommand{\labelitemii}{$\circ$}

\newcommand{\resumeSubHeadingListStart}{\begin{itemize}[leftmargin=*]}
\newcommand{\resumeSubHeadingListEnd}{\end{itemize}}
\newcommand{\resumeItemListStart}{\begin{itemize}}
\newcommand{\resumeItemListEnd}{\end{itemize}\vspace{-5pt}}
\newcommand{\awarditemListStart}{\begin{itemize}[leftmargin=*]}
\newcommand{\awarditemListEnd}{\end{itemize}\vspace{-5pt}}
    


%-------------------------------------------
%%%%%%  CV STARTS HERE  %%%%%%%%%%%%%%%%%%%%%%%%%%%%


\begin{document}

%----------HEADING-----------------
\begin{wrapfigure}{r}{2.2cm} % 'r' for right-align, and 3cm is the width of the image box
  \vspace{-\baselineskip} % Move the image up a bit to align with the top of the text
  \includegraphics[width=2.2cm, height=2.2cm]{qrcode.png}
\end{wrapfigure}

\textbf{{\Large Nikan Doosti}} \\
Email: \href{mailto:nikan.doosti@outlook.com}{nikan.doosti@outlook.com}\\
Homepage: \href{https://nikronic.com/}{https://nikronic.com}\\
Github: \href{https://github.com/Nikronic}{https://github.com/Nikronic}
% ----------------------------------

\hspace{0cm} % fix the hbox overfull caused by QRcode addition
%-----------EDUCATION-----------------
\section{Education}
 \resumeSubHeadingListStart
   \resumeSubheadingF
     {Iran University of Science and Technology (IUST)}{Tehran, Iran}
     {Master of Science in Computer Engineering - Artificial Intelligence}{Aug 2019 - Dec 2022}
      \resumeItemListStart
        \resumeItem{\textbf{Thesis:} High Resolution Neural Topology Optimization via Differentiable Physics Engine (\href{https://github.com/Nikronic/ndr}{code})}
        \resumeItem{\textbf{Defense:} Achieved \textbf{maximum score} during defense on \textit{Oct 22, 2022} with \textbf{GPA of 17.17/20.00}}
        \resumeItem{\textbf{IUST:} This university is one of the most prestigious in the country, being in \textbf{top-4} consistently.}
      \resumeItemListEnd
   \resumeSubheadingF
     {University of Guilan (UoG)}{Rasht, Iran}
     {Bachelor of Science in Computer Engineering}{Aug 2015 - Aug 2019}
     \resumeItemListStart
        \resumeItem{\textbf{Project:} Descreening and Rescreening of Halftone Images via Data-Driven Deep Learning Methods (\href{https://github.com/Nikronic/Deep-Halftoning}{code})}
        \resumeItem{\textbf{Class Rank:} Graduated \textbf{3rd} out of 55 with a \textbf{GPA of 18.64/20.00}}
      \resumeItemListEnd
 \resumeSubHeadingListEnd
%-------------------------------------------

%--------Publications------------------
\section{Publications}
\label{sec:publication}
 \resumeSubHeadingListStart
   \resumeItem{\textbf{Doosti, Nikan}, Julian Panetta, and Vahid Babaei. "Topology Optimization via Frequency Tuning of Neural Design Representations." In \textbf{Symposium on Computational Fabrication}, pp. 1-9. 2021. (\href{https://dl.acm.org/doi/abs/10.1145/3485114.3485124}{ACM}, \href{https://github.com/Nikronic/ndr}{code})}
\resumeSubHeadingListEnd
%-------------------------------------------

%--------TALKS------------------
\section{Talks}
 \resumeSubHeadingListStart
   \resumeItem{"Neural Design Representations." \textbf{Toronto Geometry Colloquium Advised by Alec Jacobson} - University of Toronto. March 4, 2022. toronto-geometry-colloquium.github.io. (Length: 10 mins., \href{https://youtu.be/FdPwG2kNv0M}{Video})} 
\resumeSubHeadingListEnd
%-------------------------------------------

%-----------RESEARCH EXPERIENCE-----------------
\section{Research Experience}
  \resumeSubHeadingListStart
    \resumeSubheading
      {Max Planck Institute for Informatics}{Saarbr{\"u}cken, Germany}
      {\textbf{Research Assistant (remote)} 
      }{Jul 2020 - Mar 2021}{\href{http://aidam.mpi-inf.mpg.de/?view=home}{Artificial Intelligence aided Design and Manufacturing Group}}
     
      \resumeItemListStart
        \resumeItem{Project Overview: \textbf{Neural Design Representation}: Novel self-supervised neural method for obtaining the optimum design showcased in Topology Optimization}
        \resumeItem{Supervisors: Supervised by \href{http://aidam.mpi-inf.mpg.de/?view=people_vahid} {\textbf{Dr. Vahid Babaei}} and collaborated with \href{https://www.julianpanetta.com/}{\textbf{Prof. Julian Panetta}} from the University of California, Davis, USA.}
        \resumeItem{Method: Utilized \textbf{physics-informed deep learning} by integrating analytical \textbf{physical simulators of PDE-constrained density-based topology optimization} into \textbf{neural fields}, enabling \textbf{sub-voxel (pixel) filtering} to control the smoothness of designs without the need for explicit smoothing filters}
        \resumeItem{Details: Applied analytical physics and classical numerical methods, such as Finite Element Methods (FEM), to simulate the objective function. Additionally, explored various neural architectures and reproduced their results, from convolutional neural networks to modern neural fields, to study the potential of deep learning for design representation.} 
        \resumeItem{Interdisciplinary Work: Successfully navigated and \textbf{mastered uncharted domains} beyond my primary field, including topics in mechanical engineering with no prior experience.}
        \resumeItem{Experimentation: Managed \textbf{large-scale experiments} by extending logging solutions such as MLFlow for model, experiment, config, and report \textbf{versioning and tracking}, integrated into Slurm workload manager on compute cluster, particularly enabling \textbf{easy follow-up near deadline}.}
        \resumeItem{Collaboration: \textbf{Shared PyTorch expertise} with group members, focusing on low-level internals and \textbf{optimizing workflows with Slurm clusters}.}
        % \resumeItem{Commitment: Dedicated \textbf{over 1500 hours} to research and development (excluding paper draft and revision), demonstrating a \textbf{strong commitment} to the project and its outcomes.}
        \resumeItem{Manuscript Development: \textbf{Prepared all figures} and contributed approximately \textbf{65\% to the manuscript}. Also, I oversaw all \textbf{administrative tasks} related to the paper's publication, including \textbf{handling revisions} and addressing \textbf{peer review feedback}.}
        \resumeItem{Outcome: Resulted in a paper published and presented at the \textbf{ACM Symposium on Computational Fabrication 2021} (see \hyperref[sec:publication]{Publications}). Additionally, I was among the few master's students whose thesis led to a publication in a highly regarded venue.}
      \resumeItemListEnd
  \resumeSubHeadingListEnd
%-------------------------------------------


%-----------TEACHING EXPERIENCE-----------------
\section{Teaching Experience}
  \resumeSubHeadingListStart
    \resumeSubheadingF
      {Head Teaching Assistant - Advanced Programming (AP)}{}
      {Supervisor: \href{https://ir.linkedin.com/in/seyed-abolghasem-mirroshandel-1a3a5950}{Dr. Ghasem Mirroshandel} - University of Guilan} 
      {Aug 2018 - Feb 2019}
  
    % taking too much space and it is not even comparable to what I did in other TAs
    % I have not even included Advanced programming and data structure helps 

    % \resumeSubheading
    %   {Teaching Assistant}{}
    %   {Algorithms Design} 
    %   {Feb 2017 - July 2017}{University of Guilan}
    %   \resumeItemListStart
    %     \resumeItem{Supervision: \href{https://www.linkedin.com/in/mojtaba-moe-shakeri-b34a2b6/}{Dr. Mojtaba Shakeri}}
    %     \resumeItem{Helped with grading of the assignments}
    %   \resumeItemListEnd
    
    \resumeSubheadingF
      {Head Teaching Assistant - Algorithms Design (AD)}{}
      {Supervisor: \href{https://www.linkedin.com/in/mojtaba-moe-shakeri-b34a2b6/}{Dr. Mojtaba Shakeri} - University of Guilan} 
      {Aug 2018 - Feb 2019}
    
    \resumeSubheadingF
      {Head Teaching Assistant - Computational Intelligence (CI)}{}
      {Supervisor: \href{https://www.linkedin.com/in/mojtaba-moe-shakeri-b34a2b6/}{Dr. Mojtaba Shakeri} - University of Guilan}
      {Feb 2018 - Jul 2018}

    \vspace{10pt}
    \textbf{Taught Java} in AP, designed and graded assignments, and \textbf{evaluated final projects}. Held \textbf{weekly Q\&A sessions}, graded assignments, and \textbf{created programming tasks} for AD and CI courses (partial \href{https://github.com/Computational-Intelligence-Fall18/Computational-Intelligence-Tutorials}{materials}).
    
  \resumeSubHeadingListEnd
%-------------------------------------------

%-----------VOLUNTARY ACTIVITIES (DEPRECATED - TOO LONG)-----------------
\section{Community and Voluntary Activities}

  \resumeSubHeadingListStart
    \resumeSubheading
      {Member}{}
      {Official forum with +50K members and authors of the PyTorch} 
      {2018 - 2022}{{Official PyTorch Forum}}
      \resumeItemListStart
        \resumeItem{A top member (15th) with 183 solutions and 566 posts (\href{https://discuss.pytorch.org/u/Nikronic/}{summary})}
        \resumeItem{Commended by \href{https://x.com/ThomasViehmann/status/1309794697049714689}{Thomas Viehmann}, author of \href{https://www.manning.com/books/deep-learning-with-pytorch}{\emph{Deep Learning with PyTorch}} book for insightful posts}
        \resumeItem{Invited twice to attend the PyTorch Developers Conference, a limited-capacity event}
      \resumeItemListEnd
  \resumeSubHeadingListEnd

  \resumeSubHeadingListStart
    \resumeSubheading
      {Organizer and Mentor}{}
      {An Open and Free Organization For Sharing Ideas, Showcasing Projects, and Mentoring Students} 
      {2019 - 2021}{\href{http://iust-projects.ir/}{IUST Projects}}
      \resumeItemListStart
        \resumeItem{Attempted to challenge the university's siloed culture through open scientific/general discussions}
        \resumeItem{Mentored junior students in preparation for going through the M.Sc thesis process, from ideation to publishing, and also job hunting.}
        \resumeItem{One of the mentorees from the start of master's, recently started working as a senior backend developer in a large software company (feedback available upon request)}
      \resumeItemListEnd
  \resumeSubHeadingListEnd

  \resumeSubHeadingListStart
    \resumeSubheading
      {Mentor and Lecturer}{}
      {An Open and Free Organization For Introducing AI and Mentorship} 
      {2018 - 2021}{Rasht School of AI}
      \resumeItemListStart
        \resumeItem{Held lectures around classical and neural-based digital image processing (\href{https://github.com/rasht-school-of-ai/Meetup-Materials}{Slides})}
        \resumeItem{Mentored one student who were interested in artificial intelligence and its applications}
      \resumeItemListEnd
  \resumeSubHeadingListEnd

  \resumeSubHeadingListStart
  \resumeSubheading
    {Teacher}{}
    {Teaching Math and Programming to Underprivileged Teenagers in Low-income Regions} 
    {2023 - 2024}{Independent work}
    \resumeItemListStart
      \resumeItem{Held weekly 1.5-hour sessions to teach math and programming}
      \resumeItem{Provided mentorship to a select few on pursuing college degrees in STEM fields}
    \resumeItemListEnd
  \resumeSubHeadingListEnd

%-------------------------------------------

%-----------VOLUNTARY ACTIVITIES-----------------
% \section{Voluntary Activities}
% \resumeSubHeadingListStart
%   \resumeSubheadingF
%     {Mentor, Lecturer, and Organizer}{}
%     {Rasht School of AI, IUST Projects, and PyTorch Forum}{2018 - 2022}
%     \resumeItemListStart
%       \resumeItem{\textbf{Lecturing:} Delivered talks on AI applications, focusing on digital image processing (\href{https://github.com/rasht-school-of-ai/Meetup-Materials}{Slides})}
%       \resumeItem{\textbf{Mentorship:} Guided students in AI and M.Sc thesis processes, from ideation to publication}
%       \resumeItem{\textbf{Organizing:} Facilitated open discussions at IUST to promote collaboration and challenge the siloed culture}
%       \resumeItem{\textbf{Community Engagement:} Active in the PyTorch Forum, ranking 15th with 183 solutions and 566 posts (\href{https://discuss.pytorch.org/u/Nikronic/}{summary}); praised for insightful contributions by \href{https://twitter.com/ThomasViehmann/status/1309794697049714689}{Thomas Viehmann}}
%     \resumeItemListEnd
% \resumeSubHeadingListEnd
%-------------------------------------------

%-----------WORK EXPERIENCE-----------------
\section{Work Experience}
  \resumeSubHeadingListStart
    \resumeSubheading
      {Self-Funded AI Venture}{Tehran, Iran}
       {\textbf{Founder and Engineer}}
      {Mar 2024 - Jul 2024}{Specializing in Automated Document Image Analysis}
      % https://arxiv.org/abs/2103.15348 => start of research for this idea
      % how I came up with the idea: when I was working as a ML specialist, I did some manual tasks of preparing data, until later they recruited a data engineer. I noticed that not all companies (especially those without IT department) are interested in revising their data pipeline, and would rather have a software-service that does that for them even proportionately. This is particularly true for developing countries where the culture of having structured data or following global standardized methods is not common at all. Also, usually in these countries the bureaucracy is massive and hence such facilitating tool could become beneficial. So I concluded that I can close the gap between the lack of having standard data pipeline and business orchestration by developing an automated document image analysis platform with no-code/low-code configuration that makes internal or external transition from unstructured denormalized documents into a more accessible and structured data, more viable. 
      % the story: many small to medium companies do not follow a structured and standard layout for their documents, hence interacting directly with other businesses (which have the same issues), make is challenging. On top of that, the variety of layouts both for internal and external interaction adds to this complexity. To resolve this challenges, many domain experts in various departments have to manually go through all these documents and use manual labor to convert each document to a format suitable for their department or overall their company when it comes from outside.
      % this massively slows down the process, wastes resources, and also cause dissatisfaction and damages if not gone through intensive manual review.
      % technical details: 1) automatically detect any structure of typed document 2) no-code/low-code configuration, such as categorization, validation against business logic, and review step-by-step via human-in-the-loop view in mind 3) integration into secondary tasks including CRM, BI Dashboards, etc.
      % failures: Unfortunately I failed due to the following reasons:
      % failure01: technology cannot change a standing culture is it is an incrementation (optimization of the current process) and not a new way (revolutionary). In my region, standardizing data hence data engineering is not appreciated.
      % failure02: one might think a good globally accepted solutions can be integrated, and built upon, but the gap is not just one implementation away. it is exponential in form of layers. If previous layer of technology doesn't exist, it is not C(L)+C(L-1), it is C(L)+C(L-1)^y.
      % failure03: governmental invasive regulations: well regulation is against innovation. I even had to pay 3x more for servers and crypto for VPN stuff to just only test my solution on cloud before getting filtered.
      % failure04: I analyzed that offices with large bureaucracy (which in a lot in my country - we are worse than Germany by far), could benefit the most. What I didn't consider was the extreme length of corruption. So practically it is preferred to have long manual processes than money saving automated one, for two reasons: 1. getting a higher budget for manual workload 2. prevention of transparency which allows easier manipulation. (e.g., there are part of the government that if you lose a paper, all data is lost) 

      % I converted some of the challenges into opportunities:
      % for instance, in case of accuracy, either we had to be 100% accurate, or the solution was undesirable. So what I did was that I made the process and the tech in a way that every step can be audited by an expert, hence making even lower accuracy solution "desirable" (now the experts only look at the flow if there are doubts.)
      \resumeItemListStart
        \resumeItem{The problem: Many small to medium companies, \textbf{lack structured data pipelines} and use their own specific layout for their documents which degrades inter-company interactions.}
        \resumeItem{Developed an automated document image analysis platform to \textbf{transform unstructured, denormalized documents into accessible, structured data}, semantically searchable.}
        \resumeItem{Created a \textbf{no-code/low-code configuration system} for easy customization and business logic validation}
        % \resumeItem{Integrated a \textbf{human-in-the-loop review process} for quality control and compliance}
        \resumeItem{Outcome and Insights: While the venture \textbf{did not achieve commercial success}, it provided valuable learnings: 
          \begin{itemize}
            \item Impact of \textbf{infrastructural inertia} toward data standardization
            % \item Complexities of \textbf{localization} of global tech solutions
            \item Effects of \textbf{regulatory environments} on innovation
            \item \textbf{Bureaucratic preferences} for transparency prevention in process management
          \end{itemize}
        }
      \resumeItemListEnd
    
    \resumeSubheading
      {Panafor}{Karaj, Iran}
       {\textbf{Full-time Machine Learning Researcher}}
      {Apr 2022 - Jan 2024}{Specializing in Data-driven Decision Making for Business Optimization}
     
      % {Here is the description:
      % 1. The problem: Tourism unlike its fun nature, has its own issues, particularly acquiring Visa. Now, we have tone of people interested in international tourism but not all can have satisfying application for a successful visa. Given huge number of calls (ignoring online platforms entirely) which is about 1000 calls per day, there was a huge overhead for the personnel to figure out whom they should prioritize as many of those calls had no or little chance of getting a visa. Also, we had limited personnel, and it was not possible to handle serious potential customers with high focus and low error in administration process without sacrificing more calls. This issue, resulted in huge costs, as a slightest error would have been fatal for the customer.
      % 2. What I did: As an artificial intelligence specialist, I figured out, instead of overwhelming the expert personnel with low likelihood customers (which only could be figured out after many minutes of consulting, we can automate the entire screening process; from communication with the customer (a), over customer data acquisition (b), to prioritizing the application of the customer (c). 
      % For step (a), a speech2text system would filter calls that have basic questions or had very weak application. Given that we had no labeled clean data, I decided that we better buy such service from third-party companies. Given the importance of privacy, I integrated an offline solution. Step (b) was more challenging as there were no solution at all available, hence I had to prepare our own data. Hence, I designed and implemented a full pipeline of automated data extraction, transformation, and preprocessing dealing with low/unlabeled data regime that reduced bad data by \%35. In the final step (c), a machine learning system (XGBoost) in combination with ExplainableAI was used to not only prioritize customers, but also explain the reason for each factor affecting the likelihood of each customer as it was necessary to communicate reasoning for such decision. 
      % 3. The results: Through the success of this project, I managed to reduce the total error of personnel by \%10, preventing the company from losing around 5.5 times of my salary. Also, I have been awarded with "Most Zealous" and "The Jedi" for working hard and being an inspiring leader respectively, resulting in two promotions and \%70 (normalized by inflation - raw \%125) increase in salary within a year.}

      \resumeItemListStart
        \resumeItem{The Problem: As the number of customers grows, assigning experts to each one becomes critical, as only a few result in contracts. An "smart operator" that can monitor all customers in real-time, identify high-value ones, and assign experts based on their fitness would help prevent wasted effort.}
        % \resumeItem{The Solution: Developed and implemented a \textbf{Data-driven AI solution} that optimized resource allocation by \textbf{prioritizing high-potential customer profiles}, significantly \textbf{reducing operational overhead and minimizing errors} in processing critical applications.}
        % \resumeItem{The Solution: 1. Screening customer input (text, voice call), 2. Finding potential via tabular machine learning methods, 3. Reporting potential and fitness to expert to human domain expert via explainableAI, 4. Experts are only engaged with the most=likely customers.}
        \resumeItem{The method:
          \begin{itemize}
            \item Screen customer input (text converted to categorical data via LLM API, voice to text via Automatic Speech Recognition)
            \item Identify high-potential customers using tabular machine learning methods, including XGBoost.
            \item Report fitness to the human domain expert via explainable AI, prototyped using Gradio.
            \item Engage experts only with the most likely customers.
          \end{itemize}
        }
        \resumeItem{Impact: Decreased personnel error by 10\%, and I established myself as \textbf{the primary person for onboarding and training} new team members.}
        \resumeItem{The \textbf{comprehensive screening process automation} (text/voice) coupled with filtering calls based on the complexity of inquiries, \textbf{reduced manual workload by 40\%.}}
        \resumeItem{Oversaw the development of a proprietary data extraction and preprocessing pipeline, resulting in a \textbf{35\% reduction in poor-quality data}.}
        % \resumeItem{Deployed \textbf{classical machine learning} models alongside \textbf{deep learning} methods, coupled with \textbf{explainable AI} techniques to prioritize profiles and provide transparent reasoning for each decision.}
        \resumeItem{Exhibited \textbf{proactive problem-solving} by \textbf{manually preparing years of "analog data" within the first 2.5 months}, a critical task which I prioritize over my role-specific duties to ensure project success.}
        \resumeItem{\textbf{Managed a 15,000-line codebase}, ensuring maintainability and performance. \textbf{Designed 7 modules, with 3 adopted by other projects}, enhancing reusability and impact.}
      \resumeItemListEnd
  \resumeSubHeadingListEnd
%-------------------------------------------

%--------Technical Skills------------
\section{Technical Skills}
  \begin{tabular}{>{\raggedright}p{4cm} @{\hskip 1cm} p{13cm}}
    Deeply Involved: & Python, PyTorch, Tensorflow, Git, Windows, Linux/Debian, MLFlow, DVC, Pandas, Sklearn, ExplainableAI, Sphinx Doc, "why you should solve a problem on top of how" \\[0.6cm]
    Have Experience With: & Docker, DevOps, CI/CD, Slurm, PostgreSQL, FastAPI, Shell Scripting, HTML/CSS, Latex
  \end{tabular}

% explanation for "why you should care": Skepticism vs Pessimism => I only resort to choosing a method based on evidence; samples, analysis or experts heuristics. This particularly makes me prune to hype-based decision making, which is particularly dominant in AI industry. So, I believe this positive skepticism which is not being pessimistic, is very important as it massively prevents unnecessary costs. And the reason I put it technical skills is that I practically use for for all levels of decision making, from writing a variable name when I code, or choosing the design of a system. So far, the fact that I demoed why AI won't work for many businesses, even though I could get paid a lot, I think is my best usage of this skill which have said probably hundred thousands of dollars.
%-------------------------------------------

%--------Research Interests------------
\section{Research Interests}
  \resumeSubHeadingListStart
    \resumeItem{Deep Learning; Physics-Informed Deep Learning}
    \vspace{-5pt}
    \resumeItem{Computer Graphics, Physics-based Simulation, and Inverse Design} 
    \vspace{-5pt}
    \resumeItem{AI for Engineering and Health; Computational Fabrication, and Drug Design}
  \resumeSubHeadingListEnd
%-------------------------------------------

%---------Awards-----------------------
\section{Awards and Certificates}
\awarditemListStart
  \awarditem{Awarded for \textbf{dedication and leadership} at Panafor}{2023}
  \awarditem{Completed \textbf{training in Workplace Professionalism}, Organizational Behavior, etc.}{2023}
  \awarditem{Accepted in M.Sc program as a \textbf{National Exceptional Talent}, with \textbf{Tuition Waiver} at IUST}{2019}
  \awarditem{\textbf{Ranked 3rd} among B.Sc graduates in Computer Engineering, with \textbf{Tuition Waiver} at the UoG}{2019}
  \awarditem{Participated in the Deep Learning Summer School at Gda\'{n}sk University of Technology}{2020}
  \awarditem{Invited twice to attend the PyTorch Developers Conference}{-}
  \awarditem{MOOC including Coursera ML and DL Specializaion (Andrew Ng), NYU DLSP (Alf Canziani), etc.}{-}

\awarditemListEnd
%-------------------------------------------

%--------Language Skills------------
\section{Language Skills}
  \resumeSubHeadingListStart
    \resumeItem{English: TOEFL 108 (Reading: 30, Listening: 27, Speaking: 23, Writing: 28)}
    \vspace{-5pt}
    \resumeItem{Persian: Native (also, Gilaki a Caspian/Daylamite language)} 
    \vspace{-5pt}
  \resumeSubHeadingListEnd
%-------------------------------------------

%-----------REFEREES-----------------
\section{Referees}
  \resumeSubHeadingListStart
    \resumeSubheading
      {Dr. Vahid Babaei (Research Scientist)}{Saarbr{\"u}cken, Germany}
       {Role: Research project supervisor}
      {\href{mailto:vbabaei@mpi-inf.mpg.de}{vbabaei@mpi-inf.mpg.de}}{Max Planck Institute for Informatics}

    \resumeSubheading
      {Prof. Julian Panetta (Assistant Professor)}{Davis, USA}
       {Role: Research project supervisor}
      {\href{mailto:jpanetta@ucdavis.edu}{jpanetta@ucdavis.edu}}{University of California, Davis}
    
    \resumeSubheading
      {Dr. Mojtaba Shakeri (Research Scientist)}{Los Angeles, USA}
       {Role: Undergraduate mentor and instructor}
      {\href{mailto:mojtaba.shakeri@gmail.com}{mojtaba.shakeri@gmail.com}}{MercuryGate (prev. Assistant Professor at University of Guilan, Rasht, Iran)}
  \resumeSubHeadingListEnd
%-------------------------------------------

\end{document}