As a Senior Software Engineer, I led the design and implementation of a custom real-time fraud classification engine deployed in the Azure cloud environment. This project required a deep understanding of existing fraud detection approaches, distributed systems, and cloud infrastructure. My role spanned the entire development lifecycle, from initial system architecture and design to hands-on coding and deployment. I collaborated closely with cross-functional teams, including data scientists, product managers, and security analysts, to ensure the system met stringent performance, scalability, and accuracy requirements. Leveraging Azure's native tools and services, we built a highly scalable, secure, and maintainable solution that integrated seamlessly with Block's existing systems.The fraud classification engine proved its value during the DIY 2024 tax season, successfully scoring every single tax return, allowing for more informed service offerings in Block's DIY process. The system demonstrated its efficacy by independently identifying all instances of fraud later confirmed by the IRS, along with additional fraudulent activities that were not previously detected by legacy systems. This achievement underscored the engine's precision and reliability, providing a significant enhancement to fraud prevention efforts.To further optimize the system, I delivered a comprehensive performance rewrite that resulted in a 650% improvement in efficiency. By analyzing bottlenecks and implementing advanced optimization techniques, I reduced response times to under 200 milliseconds, even under heavy load conditions. This improvement not only enhanced the user experience but also ensured the system's ability to handle peak traffic during critical filing periods without compromising accuracy or stability.

I led the development of a custom pilot bidding system, replacing the outdated pen-and-paper charter service previously in use. By closely collaborating with stakeholders to understand their workflows, I designed and implemented a digital platform that streamlined the pilot scheduling process, improved efficiency, and minimized errors. The new system automated bid tracking and allocation, offering pilots a user-friendly interface to participate in bidding while giving administrators centralized control over operations. This transformation not only modernized the business but also significantly reduced turnaround times for scheduling flights.A key component of the pilot bidding system was a robust auctioning engine I developed from scratch. This engine allowed pilots to place competitive bids for assignments in real time, ensuring fairness and transparency. To facilitate seamless transactions, I integrated the system with the Stripe API, enabling secure and efficient payment processing. This integration supported various payment methods and ensured compliance with industry standards, enhancing user confidence and operational reliability. By tailoring the auctioning engine to meet the unique needs of the business, I helped create a scalable solution capable of handling high transaction volumes.To support the system’s development lifecycle, I implemented containerized environments using Docker, ensuring consistency across development, staging, and production. This approach streamlined the testing and deployment processes, allowing the team to rapidly iterate and deliver features with confidence. Recognizing the need for scalability and reliability, I migrated the solution to AWS, leveraging services like EC2, S3, and RDS to optimize performance.

As a Graduate Teaching Assistant, I had the privilege of guiding three lab sessions for Programming Foundations I and II, which focused on foundational concepts in C++ programming. These sessions were instrumental in helping students transition theoretical knowledge into practical application. I designed and delivered clear, engaging demonstrations that covered topics such as control structures, object-oriented programming, and algorithm development. By fostering an interactive environment, I encouraged students to actively participate and collaborate, ensuring they gained confidence in their programming skills.In addition to leading lab sessions, I was responsible for grading homework and lab assignments. This role required me to evaluate student submissions meticulously, providing constructive feedback to help them improve their coding practices and problem-solving abilities. I also contributed to course development by creating comprehensive question banks and designing lab assignments that aligned with the course’s learning objectives. These materials were thoughtfully crafted to challenge students while reinforcing key concepts.Beyond the classroom, I maintained regular office hours to provide personalized assistance to students. This setting allowed me to address individual questions, clarify complex topics, and mentor students on debugging and code optimization techniques. On one occasion, I had the opportunity to step in and lecture to a class of 300 students. This experience honed my public speaking and instructional skills, as I prepared and delivered a lecture that was both accessible and engaging for a large audience. This blend of responsibilities enhanced my ability to adapt teaching strategies to various formats and group sizes, solidifying my commitment to fostering a supportive and enriching learning environment.

During my tenure as a Graduate Research Assistant, I played a pivotal role in a Department of Energy (DoE) sponsored project focused on developing a machine learning (ML) powered Common Vulnerabilities and Exposures (CVE) patch recommendation system. This initiative aimed to enhance cybersecurity by enabling organizations to identify and implement optimal solutions for software vulnerabilities. My contributions involved leveraging advanced ML techniques to create a system capable of intelligently recommending patches based on a wide array of factors, including vulnerability severity, software environment compatibility, and historical patch efficacy. This work not only required technical expertise but also a strong understanding of cybersecurity challenges in enterprise and government settings.One of the key challenges in this project was integrating and harmonizing data from diverse CVE providers, each of which utilized different formats and standards. To address this, I designed and implemented comprehensive process flows that unified large quantities of CVE data into a standardized format. The resulting unified dataset served as the foundation for training robust AI models, enabling the system to deliver accurate and actionable patch recommendations.In addition to data unification, I was actively involved in implementing and training the machine learning models that powered the system. These models were designed to analyze historical patching data and predict optimal solutions for new vulnerabilities. Recognizing the importance of trust and transparency in AI-driven systems, I incorporated deterministic evaluation methods to make the models’ decision-making processes fully auditable. This approach ensured that recommendations could be explained and validated, which was critical for user adoption in high-stakes cybersecurity environments.

Collaborated with a tight-knit team of developers to optimize, refactor, and migrate legacy implementations into a Symfony LTS codebase for CTTP, a state-accredited engineering company that provides training and certification services for employees and contractors of the AR Department of Transportation (ARDOT).Facilitated digitization processes and migration from a physical record keeping solution to a Symfony-powered web application, which streamlined operations and improved efficiency. As a senior member of the team, I played a key role in leading the planning and implementation of various projects, which included an online registration and payment portal, a certification and grade record keeping service, a lab inspection service for ensuring compliance with state guidelines, and an equipment calibration record management system for securely storing and access for auditing purposes. I also provided input and suggestions during meetings to solve complex problems and contributed to the deployment and server administration tasks of the system.One of my key accomplishments during my time at CTTP was containerizing the development and production environments using Docker, which significantly improved the onboarding process for new developers and deployment times. By improving the efficiency of the development and deployment environments, we were able to reduce overhead costs and accelerate the software development life cycle.Overall, my second experience at CTTP further reinforced my understanding of the full software stack, business logic, and server administration. This exposure has prepared me well and instilled me with confidence for the many future challenges and opportunities in the field of software engineering has to offer.

During my summer internship at Walmart's newly created Intelligent Automation Department on the Bentonville, AR campus, I spearheaded the design and development of an Optical Character Recognition (OCR) system tailored to digitize handwritten and printed produce receipts and invoices. Recognizing the challenges of inconsistent handwriting, diverse invoice formats, and varied printing quality, I incorporated a human moderation layer to ensure accuracy. This allowed users to validate and correct OCR outputs efficiently, resulting in a system that combined automation with human oversight for high precision.To enhance usability, I developed a user-friendly interface where moderators could review flagged items in real time. The project involved using Python and integrating popular OCR libraries like Tesseract. I also leveraged custom preprocessing techniques to improve text extraction from low-quality scans. By reducing manual data entry efforts while maintaining accuracy, the system streamlined workflows and significantly improved data reliability.As part of a project aimed at optimizing compute-intensive computer vision (CV) workflows, I designed and deployed a task management system powered by Celery and Redis. This system facilitated the distribution of computational tasks across multiple workers, effectively scaling operations beyond the limitations of a single server. The approach ensured seamless handling of high workloads, particularly for tasks like image preprocessing, feature extraction, and text recognition.The task system dynamically allocated resources based on workload demand, improving throughput and reducing latency. I also implemented robust retry mechanisms and monitoring tools to ensure task reliability and detect bottlenecks. This architecture proved essential for processing large datasets, such as batches of scanned documents, in parallel.

As a software developer for a company based in Bietigheim-Bissingen, Germany, in addition to brushing up on some german, I had the opportunity to work on a variety of projects for clients across different industries. I particularly enjoyed working on the many challenging projects that required creative problem-solving.One such project was a challenge hosted by BMW, where I contributed to an inventory system overhaul that was selected for utilization in a major manufacturing plant. Our solution consisted of a barcode scanning glove, programmable e-ink displays, and the integration with an existing MQTT system. I played a key role in the development of this system, which significantly modernized the part-request process for assembly lines and replaced a pen-and-paper solution. The project presented many unique challenges, such as only being provided limited, API-like details of the internal MQTT system at BMW and the need to integrate the custom, in-house hardware with a control system whilst simulatenously on simulatating a MQTT network.In another project, I contributed to a seed grant sponsored by the EU to enable public transit for rural towns in Germany. This project, dubbed "der bürgerbus" (the citizen bus), required a custom routing solution to allow for dynamic time and location scheduling of stops. I extended an open source routing engine "Project OSRM" to handle the dynamic constraints and developed a web user interface that allowed operators to view and modify proposed routing solutions in a intuiative manner. This was one of the most complex problems I have had the pleasure of programming.Largely, these experiences exposed me to a range of unique challenges and provided me with valuable experience in solving complex problems and developing creative solutions. They also taught me the importance of collaboration and communication in achieving project success.

Collaborated with a small team of developers to optimize, refactor, and migrate legacy implementations into a Symfony LTS codebase for CTTP, a state-accredited engineering company that provides training and certification services for employees and contractors of the AR Department of Transportation (ARDOT).Facilitated digitization processes and migration from a physical record keeping solution to a Symfony-powered web application, which streamlined operations and improved efficiency. Played a key role in improving the efficiency and productivity of the company by contributing to the development of several applications that were used by both internal and external stakeholders. These applications include, but are not limited to, an online registration and payment portal, a certification and grade record keeping service, a lab inspection service for ensuring compliance with state guidelines, and an equipment calibration record management system for securely storing and access for auditing purposes. These contributions helped the company to better serve its clients, reduce administrative overhead, and improve accuracy and timeliness of its operations.Altogether, my experience at CTTP was instrumental in shaping my career as a software engineer. By working on a range of projects and collaborating with a talented team, I gained valuable experience in all aspects of the software development life cycle. This exposure to the full software stack and business logic has prepared me well for future challenges and opportunities.