Hi, I'm Furqan Arshad.
A
Self-driven, quick starter, passionate programmer with a curious mind who enjoys solving complex and challenging real-world problems.
About
Furqan Arshad is a computer scientist with expertise in neuroscience, medical signal processing, AR/VR, and embedded systems. He has a track record of innovation and extensive experience in the design and development of cutting-edge technologies, as well as leading multidisciplinary teams of data scientists and software engineers. He is currently working at the University of California, Davis, where he leverages his expertise to innovate in the field of brain-machine interfaces and wearables. Previously, he worked at META Innovation Lab, where he contributed to augmented and virtual reality projects and led pioneering efforts in early migraine prediction using wearable time series sensor data. He holds a Master’s degree in Computer Science from UC Davis, focusing on biological decoding and neural interfaces. His research and professional contributions have been recognized with publications and industry accolades.
Furqan is currently seeking full-time roles in Data Science or Software Engineering where he can apply his skills and contribute to innovative projects.
Experience
- Developed and deployed end-to-end scalable neural decoding software to parse, classify, and analyze neural data from M1 and S1 brain regions using signal processing, dimensionality reduction, and classical ML techniques (Neural Networks, PCA, PSTH, Linear Dynamic Systems). Achieved 90% accuracy in predicting pull duration, significantly aiding spinal cord injury patients.
- Led neural data analysis on Monkeys during various behavioral tasks, including joystick pulls, Vonfrey and center-out games, providing insights into neural responses pre- and post-spinal cord injury.
- Developed a circuit board for synchronizing control between a robot, sensors, neural stimulation devices, photodiodes, and Plexon biological signal recording hardware using an embedded device.
- Managed and tracked multiple recording sessions across various tasks and animals, ensuring data integrity, resolving issues, and authoring detailed DARPA reports to present findings using A/B testing to both technical and nontechnical audiences.
- Researched AI trends and worked with technical and non-technical specialists to automate research processes. Created and documented interactive reports with Power BI, Seaborn, and Matplotlib, presenting insights to stakeholders using custom KPIs
- Collaborated on a DARPA project to test a blood monitoring sensor for hemodynamic management, using Python for signal processing, alignment, peak extraction, and calibration, achieving 99\% benchtop accuracy
- Tools: Python, MATLAB, Neuroexplorer
- Collaborated on the MigraineAid project, integrating multi-modal data from weather, sleep, and stress inputs, along with digital biomarkers from a smartwatch, to build a comprehensive data library.
- Built an early migraine prediction system using wearable time series sensor data combining patient diary data (Firebase) and weather data, achieving 87% accuracy. Utilized TensorFlow, LSTM, CNN, AWS for deployment, and Apache Spark for data processing.
- Pioneered end-to-end product life cycle of firefighter training system: developing realistic fire scenarios on Unity3d using C#, managing a cross-functional team of software developers and 2 UX/UI designers.
- Crafted a cutting-edge VR training module using Unity 3D, C#, and VIVE/Quest for sexual harassment awareness in Pakistan, significantly boosting bystander intervention willingness and understanding, as evidenced by comprehensive A/B testing. Tools: Python, Flask, OpenCV, Keras, Tensorflow, PyTorch
- Directed the creation of an advanced remote assistance application for hololens using Unity3D, webRTC, and Arcore SDK to reshape remote assistance in technical troubleshooting scenarios, achieving 75% accuracy in pinpointing machine parts.
- Orchestrated a drone control application for Microsoft HoloLens and ODG R7 using voice commands and hand gestures, reducing control errors by 40%.
- Tools: Unity3d, C#, Python
- Engineered end-to-end AR/VR applications (AR Furniture, VR Endoscopy, AR Shopping) using Unity3D, ARCore, and ARKit for headsets (Vive, Oculus, HoloLens). Partnered with UX team to analyze user behavior, enhancing user engagement by 300%.
- Tools: Unity3d, C#, Python
Projects
Predicting migraine attacks with 86% accuracy, 24 hours in advance using multi-modal data.
- Tools: Python, Machine Learning, Smartwatch Data Integration, Digital Diary Analysis
- Developed software infrastructure for data collection and processing.
- Implemented machine learning algorithms for pattern discovery and migraine prediction.
- Achieved 86% accuracy in predicting migraine episodes 24 hours in advance.
A mobile application designed for hepatitis awareness and scheduling appointments on Android and iOS.
- Tools: Java, Kotlin, Swift, Firebase, Android Studio, Xcode
- Provides information on hepatitis symptoms, causes, prevention, and cure.
- Allows users to track appointments and receive medicine reminders.
- Helps users assess their risk of hepatitis and encourages timely diagnosis.
- Available in Urdu, English, and Roman Urdu to cater to diverse user demographics.
- Includes audio explanations for better understanding, especially for users with literacy limitations.
Developing and Testing Full-Implantable Blood Pressure Monitoring Sensor
- Tools: Python, Signal Processing, Machine Learning, A/B testing, Azure, Power BI
- Collaborated on a DARPA project to design and test a medical device for spinal cord injury patients, improving performance by 200% through real-time signal processing and calibration, and achieving 99% benchtop and 80% live subject accuracy.
Analyzing gender disparities in high-quality research in Pakistan from 1947 to 2020.
- Tools: Namsor API, Data Preprocessing, Statistical Analysis, Visualization
- Determined the gender of authors, providing critical insights into gender disparity in authorship.
- Performed extensive data preprocessing, statistical analysis, and visualization to present the research landscape effectively.
- Analyzed a Scopus dataset of 173,950 papers revealing women account for 19.44% of authors.
- Significant increase in female authorship from 30 in the 1950-1960 decade to 38,610 in the 2010-2020 decade.
Evaluated animated videos' effectiveness on bystander intervention for sexual harassment.
Developed a state retention solution on the STM32 L152RE board using hashing, achieving a 60% improvement.
Virtual reality training system for firefighters using Unity 3D.
Improved neural data analysis for Brain Machine Interface, achieving a 50% boost in predicting joystick gaming behavior.
- Tools: MATLAB, Python, NeuroExplorer, Plexon, Photodiode Circuit Design
- Achieved a 15% increase in data processing speed and improved user experience by leading the development of real-time neural signal decoding and processing software for BCI.
- Reduced data acquisition errors by 90% through precision photodiode circuit design for accurate timestamping in direct neural system interfacing.
Skills
Languages and Databases
Libraries
Frameworks
Other
Education
University of California, Davis
Davis, USA
Degree: Master of Science in Computer Science
Dates: Sep 2021 – Jun 2023
CGPA: 3.88/4.0
Lahore University of Management Sciences
Lahore, Pakistan
Degree: Master of Science in Computer Science
Dates: Mar 2015 – Jun 2017
University of Management and Technology
Lahore, Pakistan
Degree: Bachelor of Science in Electrical Engineering
Dates: Mar 2010 – Dec 2014