Matthew C. Dewey
Verified Expert in Engineering
Software Architect and Developer
Matthew is a seasoned team leader, developer, and architect with 20+ years in C, C++, and Python, focusing on video production and backup storage. An expert in embedded systems, IoT, real-time and performance-driven systems, he excels across the software stack. Having led remote teams of two to 50+ members, Matthew boasts a deep knowledge of storage systems, fault-tolerance, and cloud development, relishing challenges in development, architecture, and team management.
Python, Linux, C++, C
The most amazing...
...machine learning implementation I've delivered to address key customer concerns increased the system's throughput by 30%, doubling its maximum capacity.
- Trained a deep neural network to be used in industrial inspection.
- Implemented serverless front and back-end processing in AWS to deploy a complete solution.
- Enhanced time-lapse camera system video rendering using deep neural network pose estimations.
- Executed a multiprocessing pipeline to increase the speed and resolution of motion detection.
Data Engineering Manager
- Led a team of nine direct reports across multiple scrum teams in an AWS environment with multiple deploys daily.
- Spearheaded the team in establishing monitoring and alerting systems for deployed services, and conducted post-incident reviews and remediation to ensure optimal performance.
- Refactored the system event pipelines by transitioning the Logstash application from an EC2 instance to a container orchestration platform using HashiCorp Nomad.
- Hired and onboarded 11 software and data engineers.
- Led the architecture and development team for a deduplication appliance deployed as a hardware appliance, virtual machine, or in a public cloud environment (AWS).
- Coordinated communications across functional and business units.
- Initiated and implemented a performance modeling project, utilizing the obtained results to design improvements for the existing product. These changes enhanced profit margins and simultaneously increased overall system performance.
- Guided the team in understanding new features, analyzing competitive factors, and providing design and implementation recommendations from the initial concept to deployment and customer support.
- Designed a hyper-converged platform for diverse applications, including multiple product lines, deduplication appliances, video surveillance, and editing systems, as well as block, object, file, and hyper-converged storage solutions.
- Refined the company's development processes by adopting more agile, software-centric, and lean methodologies. This change resulted in more predictable development timelines and significantly improved overall productivity.
- Initiated and led multiple development projects and streamlined company processes that decreased service rates, improved tracking of customer issues, and reduced service costs, contributing to the highest margin service revenue for the company.
Lead Software Architect
- Led the team in designing and developing a clustered, heterogeneous file system accessible from Linux, Windows, and Sun Microsystems (Solaris) operating systems.
- Architected and implemented a coherent client file system cache that ensures data consistency and seamless access. The solution includes advanced failover and recovery mechanisms for enhanced system stability and reliability.
- Acted as the CTO, significantly contributing to the eventual successful acquisition of the company.
- Led the architectural design and implementation team for object-oriented RAID software, ensuring robust and efficient system performance.
- Oversaw the development of several fault-tolerant RAID read and write cache systems as the lead software engineer.
- Served as the lead software engineer for real-time subsystem monitoring while developing domain-specific language to streamline the implementation of time-dependent condition monitoring and fault-tolerant storage system responses.
- Created firmware architecture for whole house audio system.
- Designed the digital hardware for this system, including the main controller and the remote control panels.
- Implemented real-time decoding of infra-red control signals.
- Designed and implemented a token-passing network, routing and addressing across multiple controllers and remote panel networks.
- Designed and implemented a real-time multi-threading system for use in storage controllers.
- Led team implementing firmware for caching storage controller.
- Implemented the caching layer, including zero latency read and minimum latency write algorithms and the associated micro-code for a proprietary microsequencer.
- Led design and implementation of human interface software for high-frequency data modem shipboard communications.
- Analyzed existing implementation to ensure minimal operator retraining was required when replacing existing systems.
- Implemented discrete Fourier transform, BCH decoder, and other signals analysis in microcode for custom bit-slice processor while meeting hard real-time requirements.
DXi Backup Appliancehttps://www.quantum.com/en/products/backup-appliances/
F-Series NVME Storagehttps://www.quantum.com/f-series
Over three months, we conducted an in-depth analysis of the product requirements and investigated both NUMA and IO architectures of system motherboards for the project, which allowed us to identify suitable vendors capable of meeting the necessary specifications. Once a vendor was selected, we troubleshot system throughput issues by tracing them to their origin in a Linux kernel driver. We then carried out a thorough analysis of the driver structure to pinpoint bottlenecks. By proposing and implementing strategic configuration changes, we successfully bypassed these bottlenecks. Furthermore, we assessed the performance of the proprietary filesystem client, identifying bottlenecks in the IO dispatch of the client. We collaborated with the filesystem developer to address and resolve these issues.
Assembler, Python 3, C++, C, Python, Bash Script, Embedded C++, Microcode, APL, Pascal, Snowflake
NumPy, OpenCV, AWS Amplify, ZeroMQ, Keras
Linux, Embedded Linux, Docker, Raspberry Pi, Raspbian, AWS Lambda, Raspberry Pi 3 GPIO, Kubernetes, AWS IoT, VxWorks
Linux File Systems, Amazon S3 (AWS S3), Datadog, Windows File Systems
Linux Kernel, Storage, File Systems, RAID, Blob Storage, Object Storage, Deduplication, Distributed File Systems, Embedded Systems, Leadership, Performance, Debugging, Embedded Software, Firmware, Multithreading, Hardware, Algorithms, Technical Leadership, Software Engineering, SaaS, Yacc, Source Code Lexing, Digital Electronics, Digital Signal Processing, Engineering Management, Vue 3, CVAT, Machine Learning, Internet of Things (IoT)
Helm, Mathematica, KVM/Qemu, Amazon Cognito, Jira
Bachelor's Degree in Computer Engineering
Syracuse University - Syracuse, NY, USA