Utsab Khakurel

Æ 202-***-**** [ ad2l15@r.postjobfree.com LinkedIn GitHub Hyattsville, MD Education

• Howard University Washington D.C.

PhD, Computer Science, GPA: 3.95 Jan. 2021 – May 2024 MS, Computer Science, GPA: 3.90 Aug. 2018 – Dec. 2019 BS, Computer Science, GPA: 3.65 Aug. 2014 – May 2018 Technical Skills

• Programming: Python, C++, Java, SQL, Typescript

• Libraries: Tensorflow, Pandas, Keras, NumPy, Scikit-learn, PyTorch, Matplotlib

• Tools: Apache Kafka, SQLite, PostgreSQL, SQL Server, AWS, Django, REST API, Git, Angular5, Adobe XD, Jira Research Experience

•

Howard University Washington, D.C.

Research Assistant, AI/ML Jan 2021 - Present

Assessed contextual bias in binary image classification, yielding a 5% model generalization improvement post-augmentation using CNN models ResNet-50 and InceptionV3 trained on Men/Women datasets.

Conceptualized and designed a real-time physical threat detection framework for edge data, blending computer vision techniques with a modified YOLOv5-s backbone and online learning.

Developed the ’Detect and Alleviate Bias’ (DAB) framework, achieving a 21% weighted reduction in bias in the ML pipeline, with modules for data preprocessing, model optimization, situation testing, and bias minimization. Work Experience

•

StretchYo Remote

Machine Learning Intern Aug 2023 - Present

Architecting a hybrid system that seamlessly integrates item-based collaborative filtering and content-based recommendation, utilizing both explicit survey data and implicit user activity data for precise personalized recommendations.

•

Quiet Platforms Remote

Software Engineer Intern June 2022 - Aug 2022

Designed database and implemented locally via SQLite for multi-hop route planning algorithm in shipment engine.

Developed shortest-path shipment planner solution for cost-effective package delivery within SLA.

Achieved 99% runtime optimization via Indexed tables and additional 5% improvement with hashmap substitution.

•

NASA Goddard Space Flight Center Greenbelt, MD

Software Engineer Co-Op/Intern Oct 2018 - Dec 2018

Designed Python mechanism for diverse satellite data processing with Pandas.

Implemented and optimized altimetric height profile searches for multiple satellites reducing search time by 25%.

Developed an internal Python tool for aggregating and processing satellite data across various formats, including ASCII, binary, and SPICE reference frames.

•

Stanford Linear Accelerator Center (SLAC) Menlo Park, CA LCLS Summer Intern June 2017 - Aug 2017

Achieved a close R-free value of 0.4 when predicting atom structures from noisy images and comparing them to the original protein structure.

Utilized Python libraries like (NumPy, SciPy, RegEx) to parse stream files into the HDF5 format.

Integrated Python scripts into an automation pipeline, eliminating defective pixels from detector pad images. Projects

•

FastChain Washington, D.C.

Master’s Thesis Jan 2019 - Dec 2019

Spearheaded the development of an IoT device simulator tailored for battlefield scenarios in NS3, pioneering a resource optimization strategy through C++ backtracking, bolstered by Blockchain for secure information retrieval

(Received Best Student Paper Award, IEEE ICII, 2019).

Contact this candidate