dobsonddev Goto Github PK

🔭 I’m love building real-world solutions through generalized, scalable systems. I enjoy collaborative, dynamic work environments that are focused on building healthy, scalable, generalized systems with interesting real world applications - whether it be full-stack webapp application development, cloud/data engineering, or an in-house computer vision model being implemented into a jetson nano-equipped drone fleet to solve inefficient inventory tracking processes.

Techs: I have professional and personal full-stack experience across a wide array of modern technologies:

• Angular 16/Apollo/GraphQL<--->Spring/MongoDB/AWS (1yr @ Savvas Learning Company)
• React/Axios<---->Flask/OpenCV/In-house ML models/Supabase<----->DJI Matrice Drone/DJI Mobile SDK (8mo @ Mercedes Benz USA + IFA Group)
• React/Axios<-->Django/PSQL/AWS (6 months @ Vision Software Solutions)
• Next.js/TailwindCSS<---->Prisma/PostgreSQL/AWS (Personal and School Projects, I really love this stack)
• Flutter<--->Supabase (Personal and School Projects)
• Expo/React Native<--->Firebase (Personal Project)

Most recently, I'm leading a research team to build an automated drone warehouse inventory scanning system for IFA Group Charleston's main warehouse, a major global automotive-parts manufacturer. This is is a fully end-to-end system, from a user-friendly React UI for weekly flight scheduling, reporting, and maintenance to the backend + drone system that is running multi-layered decision making inside of a 3D graph data structure + an A* Algorithm with custom hueristics relevant to our application (batteryCost, safetyCost, obstacleAvoidanceCost etc...). At the lowest level in our drone system, we have computer vision models aiding with granular adjustments within our 3D "world" and performing the actual inventory scanning, resulting in 99% accurate barcode scanning across thousands of inventory locations. This lowest level also confirms "expected" location within our 3D model with its "actual" by calculating the trigonometry between two sets of fiducial markers. This additional layer of localization ensures the drone system is robust and determinant. There is a tremendous amount of intermediary business logic and data translation between the subsystems, so it's been a massive learning experience to design, implement, and integrate all of them to communicate and flow in harmony within the greater system.

👯 I’m always looking to collaborate and contribute to interesting projects!

📫 How to reach me: [email protected]

⚡ Fun facts: I love peppers and fly-fishing and am the only recorded person the USA to get staph infection from a guinea pig (suuuuuuper fun 10/10 recommend it)