From May 2024 to May 2025 as a Senior Development Engineer I led the connectivity (LTE/NR, Wi-Fi, and Bluetooth) engineering team’s effort to build the next generation license plate reader for Flock Safety with Linux/Android as the target OS. I was responsible for all (Linux/Android) connectivity device drivers, Android system daemons, startup scripting and SE Linux rules.

I was responsible for porting the reference C++ Android Radio Interface Layer Daemon (rild) for LTE/NR network connectivity control for use with Quectel modem modules. This effort included determining correct AT modem commands for specific Quectel modems and message handling from the Android Telephony layer to provide signal strength, statistical and configuration information to the company’s Kotlin “Phone Home” Android application.

Additionally I maintained the modem connectivity sections of the Kotlin “Phone Home” Android application to reflect changes to the rild. This allowed installers to use the information as a site survey when determining device placement. The “Phone Home” application’s information is also sent to the Flock Cloud back-end for viewing on the Flock Web Portal.

I also created and maintained the C++ Subscriber Identity Module Daemon (simd). The simd monitors the insertion status of the SIM card in the license plate reader device. The simd is responsible for switching between the external physical SIM card and the internal eSIM module. The simd monitors the insertion status based on a GPIO interrupt managed by the “sim-detect” device driver. The simd also allows selection from among the various installed provider profiles on the eSIM module.

From September 2022 to May 2024 I led the embedded software engineering team at Deer Management Systems (DMS). As a Senior Research and Development Engineer I led the company’s effort to build an internal engineering team in order to reduce dependency on external vendors and maintain control of intellectual property.

I led the firmware development for Tactacam’s first security camera. The camera is based on the Ingenic T31 (MIPS CPU) SoC and GC2064 image sensor. Main camera firmware was hosted on embedded Linux and implemented in C, C++, and bash scripting. The platform also contains an ARM MCU that controls GPIO interrupts from external devices and external Bluetooth communications. The MCU firmware was implemented as a bare-metal application. The camera operation is activated by a PIR (passive infrared) motion sensor or button press. Upon activation the camera captures images (JPEG) and/or video (MP4 encoded h.265), then uploads the results to AWS using a Quectel Cellular module or a SiliconLabs WiFi radio. I led the team in image capture, AAC and MP4 encoding and AWS functionality (S3 and MQTT communications). I was also solely responsible for the firmware optimization effort. I improved the firmware/camera performance from capturing (and uploading to AWS) 1850 images per battery charge to more than 5000 images. I also designed and created the end of line manufacturing tests for product delivery.

I also worked with multiple external vendors to create the company’s first bird feeder camera using the Ingenic T31 SoC and the next generation of trail and security cameras using the Rockchip RV1106 SoC. I participated in the initial hardware design efforts evaluating and recommending potential parts. I also provided reference firmware in an effort to train the external engineers in multiple areas (image capture, MP4 encoding, communications (LTE, Wi-Fi, Bluetooth) and AWS. I also led the internal hardware and firmware validation efforts.

I was also responsible for testing and validating new firmware releases provided by external vendors for the company’s existing (1st and 2nd generation) trail cameras.

From October 2021 to July 2022 I was a member of Google’s ‘Android SDK Research’ team (part of the Google Play Protect project). At Google I was tasked with performing security analyses on Android SDKs. This effort consisted of creating test/sample applications based on the SDK under review, decompiling the resulting APK using tools such as apktool and jadx (and Google’s internal proprietary tools), and then validating any security issues discovered using Google’s static and dynamic analysis tools. The test applications were required to exercise as much of the target SDK as possible. I was responsible for all aspects of application development. I performed all application design, implementation, and testing. A complete development life cycle was required to ensure that a true analysis of the SDK could be performed. The SDKs reviewed included advertising, analytics, remote file access, a/b testing, language translation, multiple Google Firebase SDKs and more. The test applications were created using Android Studio and Java.

I also created a Smali code instrumentation system of Python scripts (targeting Linux and MacOS host platforms). This system decompiled the target APK, extracted the Smali source code for the SDK, created a new Smali project, added debugging instrumentation as desired, compiled the project into a JAR artifact and then copied the JAR file to the correct file system location for the original APK’s linking step. This instrumentation allowed dynamic analysis of code coverage, call-stacks and various data dumps.

From May 2017 to September 2021, May 2014 to December 2014 and January 2003 to May 2006 I provided IT and software engineering consulting services to several small and medium sized West Coast companies. I was responsible for the software development needs of our IT service clients. This included diverse projects such as Android app development (Java), Android app reverse engineering and modification (Smali), embedded real-time device drivers and controllers, OTA update daemons for embedded processors, C/C++/C# Windows desktop applications, custom Android builds and various big data applications (C and Python). I have also managed several data center build and remodel projects where I was responsible for all cabling, server conditioning and hardware integration.

I also contributed to and managed several full stack LAMP projects. I provided both back-end (server side PHP scripts) and front-end (browser based JavaScript apps) solutions for multiple web applications.

From January 2017 to September 2017 I was responsible for all low-level Linux embedded systems. I created a custom secondary boot loader (SBL1) for the Qualcomm Snapdragon 410 SoC (msm8016) to initialize the SoloShot base, load various (non-Linux specific) boot images and continue the boot process. The SBL1 was built on Ubuntu Linux and cross-compiled for ARM using the ARM RVCT compiler suite. I implemented an ExFat kernel file system driver for Linux Android and heavily modified the vold (Volume Daemon) to take advantage of multiple SD Card file systems during auto-mounting. I also implemented an I2C Linux device driver for the Toshiba TC358743XBG HDMI to CSI (MIPI Camera Serial Interface) bridge chip. This included modifying the msm kernel audio subsystem to allow the correct I2S audio format. I also modified the Little Kernel (apps bootloader) to initialize the QuickLogic BX5BxA DSI (MIPI Display Serial Interface) to RGB bridge chip. I also created several Android Apps (Java and C++) for testing of the above kernel device drivers.

From June 2016 to December 2016 I provided bug fixes and functional enhancements for the company’s payment agent software running on a variety of Verifone payment terminals. All target development was done for an embedded security enhanced Linux based on the Buildroot project. Development was done in C/C++ and cross-compiled using GNU tools. I also provided HTML5 user interface bug fixes and enhancements (all Verifone payment terminals use HTML5 for user interfaces).

From September 2015 to June 2016 I contributed to the company’s In-flight Entertainment system by creating the Service Discovery and Event Server middle-ware applications. Together these applications allow a seat-back or other passenger device to query for currently running services and then to be notified when any services change. These applications were developed in C/C++ for Linux using a REST API and FastCGI through an NGINX server.

I also created a series of prototype/proof-of-concept apps for Android using JNI to provide fast access to binary files containing media information. I created the Android user interface (Java), the JNI shared library (C/C++) and the binary file format that allowed C structures to be serialized and deserialized directly without any additional parsing.

From December 2014 to July 2015 I positively contributed to multiple projects. I provided several bug fixes and user interface enhancements to GoPro’s Android and iOS camera control apps. I was also responsible for the creation of a test framework used to validate the external (WiFi and Bluetooth) interfaces to several models of GoPro cameras (including four new cameras currently in development). This included all aspects of camera control as well as validating JPEG and MP4 creation and validating GoPro’s MPEG-2 live streaming protocol (live streaming from camera to the GoPro mobile app). The test framework was compatible with Windows 7+, Mac OSX, and Linux. All software was written in C/C++ and included the integration of the Python 3.x interpreter for easy test creation/scripting.

From February 2010 to May 2014 I created the Target Image Simulation Environment (TISE) that allows direct testing of the audio, voice, video and sensor functions of the Hexagon DSP (QDSP6) in a Windows 7 desktop environment. My software (running in the HLOS) provides the communication infrastructure (via shared memory) required for DSP control. It also models several other CPU cores, external peripherals, ASIC functions and memory buses that are present on the Qualcomm Snapdragon product line. My software is used internally by Qualcomm developers and is also released externally to OEM and ISV customers (for use in development and testing of new products). It is also distributed as part of the Hexagon SDK program.

I also provided direct technical support to OEM and ISV customers and bug fixes for the Qualcomm board support package, audio, voice, video, and sensors teams. This support included all levels of JTAG and Trace32 training and scripting as well as generic embedded software optimization in C and C++ for all customers on the Hexagon DSP and Snapdragon SoC.

Also as part of the TISE project, I created a diagnostic tool that communicates directly with an Android or Windows phone. This tool collects diagnostic and performance data from all CPU cores on a Snapdragon SoC via a USB connection (Android) or via a TCP/IP connection (Windows and TISE). The diagnostic and performance data is then displayed for human monitoring and logged for offline processing / analysis.

I also created a test framework for Qualcomm’s Hexagon DSP (QDSP6). The software consisted of command-line applications designed to run on the Linux Android platform. These applications exercised all audio and voice functionalities of the Hexagon DSP in an HLOS agnostic way. I also created the Linux device drivers used to control the audio and voice features of the Hexagon DSP (QDSP6) and an Android UI based test launcher app. The Android app first discovered the capabilities of the device and then presented the user with a list of applicable tests. The app then collected all test output and logged it for off-line processing. I maintained this app for the 8650 and 8660 versions of the Snapdragon SoC.