John Rachwan

Experience

• 

  Dar

  Jun 2017 - Aug 2017

  Data Engineer

  • Implemented an Automated Phishing Program currently utilized by the company to spread security awareness.• Hybrid Front-end development of augmented reality application using ionic and AngularJS.• Taming big data using hadoop and all of its immersive and useful tools.• Developed project presentation skills while presenting for the head of software department at DAR.

• 

  APGAR

  Jul 2018 - Jan 2019

  • Developed data integration projects for pharmaceutical powerhouse IPSEN using informatica cloud.• Developed many real time processes using ICRT.• Developed Active Surveillance software for IPSEN that monitors the health of secure agents and many other parameters using Python and REST API

  • Data Consultant

    Sept 2018 - Jan 2019

  • Data Consultant (Intern)

    Jul 2018 - Aug 2018
• 

  BMW Group

  Feb 2019 - Jul 2019

  R&D, Computer Vision

  Bachelor of Computer Engineering thesis:1- 3D Labeling Tool- Implemented nonrigd registration algorithm Thin Plate Spline Robust Point Matching (TPS-RPM) in Python and optimized it for parallel computation- Smart minimum bounding rectangle calculation- Implemented the unsupervised outlier detection algorithm rPCA to avoid 3D to 2D projection mishaps- Developed an alternate way of labeling in the rare cases where the mesh does not exist by rendering the problem into a non-linear optimization problem and solving using dogbox and Newton-Raphson methods - Implemented snapping feature where the labeling object will be snapped into its correct position without any user intervention (zculling + Support Vector Registration)2- Data Preperation- Developed a workflow for labeling large pointclouds, by performing surface reconstruction using Screened Poisson Reconstruction algorithm- Abstracted the workflow in a docker container to make it simple to any user- To break the hardware limit set on the reconstruction resolution, I implemented a separate workflow that cuts the point cloud into multiple parts reconstructs each part separately and then merges them back together into a large mesh Weniger anzeigen

• 

  ETH Zürich

  Jan 2020 - Jul 2020

  Blockchain Researcher

  Working with ETH's COSS chair (https://coss.ethz.ch/) to implement a Blockchain-Enabled Verification System for the Finance 4.0 Project.

• 

  Design AI

  Sept 2020 - Jan 2022

  • Team Lead responsible of a team of 6 people• Working on the full reimplementation of DeepMind’s AlphaStar• Transferring AlphaStar to Machine Planning on Command Modern Operations• Won the EDA prize for most innovative defence project on the European scale

  • Artificial Intelligence Engineer (Team Lead)

    Nov 2020 - Jan 2022

  • Artificial Intelligence Engineer

    Sept 2020 - Nov 2020
• 

  Technische Universität München

  Nov 2020 - Apr 2021

  Teaching master-level class, on advanced data analysis and storage methods such as Hashing, Advance SQL Queries, MapReduce, C++ Optimization Techniques, OS Memory Management, NoSQL Databases, Distributed Computing. • Developed novel Dense Video Captioning model• First Dense Video Captioning model that leverages Online Learning• Beats state of the art on all metrics

  • Teaching Assistant

    Nov 2020 - Apr 2021

  • Data Scientist @ TUM Data Innovation Lab

    Nov 2020 - Feb 2021
• 

  Technische Universität München

  Sept 2022 - Aug 2023

  Machine Learning Engineer

  Working on Efficient Machine Learning Research

• 

  Pruna AI

  Jun 2023 - now

  Co-Founder & CTO

  We empower organisations that deploy AI to maximise performance by auto-compressing their models. Reduce your AI costs, times, memory & carbon emissions by 2-to-10x- Funded via Exist Government Program- Winners of AI+Munich Startup Competition- Part of Nvidia Inception Program