Tek Horizons, Sept 2017 - Present
Founder at Tek Horizons
- A consultancy founded on deliverying professional backend developer services
- Fractional CTO/Technical Leadership
- MVP Data fusion algorithm (voting)
- Audio signal processing and deploying ML detection techniques
- Lookahead traversal architecture for optimizing day planning with background build system + notifications
- API, Socket Servers, Realtime Data
- Custom applications: native and mobile web app backends, social data driven itinerary planning, optimal street routing, travel time matrix, real time geo spatial engines, video processing, custom apns/fcm push services
- Documented, tested, and benchmarked deliverables
Dorsia, May 2023 - Feb 2024
ML/Backend Engineer at Dorsia
- Designed and Deployed 3D Secure Asynchronous Payment Processing System
- Reworked onboarding logic and flow to support temporary guest passes
- Created weekly sanitized version of primary database
- Generated reservation tracking system and UI to support frictionless connections
Welcome, Jan 2020 - May 2023
Tech Co-founder at Welcome
- ML Realtime collaborative filter architecture
- ML Realtime and Interval model training and http prediction services
- ML Collaborative filter/Recommendation Engines
- Backend work supporting our Data, API & CMS
- Custom "Best Of" Scoring and Processing for Vetted
- Image processing and video simulation service
Citizen (sp0n llc), Nov 2015 - Sept 2017
Technical Co-founder at Citizen (sp0n)
- Architect: all backend services
- API (express), JWT auth, fully documented and covered by tests
- Custom data store: redis cache backed, replicated local application. 3D lat,lon,time queries
- Implemented and deployed live video streaming platform (nodejs & service workers): variations using Wowza, Red5, and nginx-rtmp
- Custom cross platform (iOS/Android) geo push architecture
- Live stream push workers
- Video post processing workers
- Deployment (Heroku + AWS) and management (loggly, custom + heroku alerts, node ping)
- Testing, benchmarking, load testing
Vimeo, April 2014 - Oct 2015
Principal Developer: for Cameo product (Fast Society)
- Designed and implemented structured theme scripts, gl shader effects, and font styling for Cameo video rendering (swift/opengl)
- Closely worked with AVComposition, slow motion video segments, audio mixing, and gl transform handling for arbitrary source video
- Designed and developed background video uploader for Cameo mobile app (swift/objc) with data validation to Google Cloud Storage
- Optimized locally exported video, tuning SDAVAssetExportSession and worked with server team for high quality streams
- Troubleshooting issues across the Cameo mobile app (swift/fabric)
- Crafted tools essential to our backend video processing architecture (c++11/libffmpeg)
- Architected and implemented our video rendering engine c++11/opencl/opengl, visual effects in opencl/gl, and structured script engine (c++11)
- Identify and resolve issues in our cloud infrastructure (node/c++) and dedicated hardware. Regular use of Kibana/Elastic Search/Syslog for debugging distributed processing
Cameo (Fast Society), April 2012 - March 2014
Chief of Research and Development: Algorithm and software development to support video processing toolchain.
- Build out backend video processing architecture (tools, dataflow, encoding)
- Optimize video processing pipeline and instance utilization
- Stability testing multiple components
- Ported JavaScript animation library to C++ opengl rendering engine
- Automated new visual animations/effects
- Aid in configuring and deploying our cloud infrastructure (basics, build error/dependency, migrated from fabric to puppet)
- Everything else our startup required
Photon Research Associates, Jan. 1996 - April 2012
Senior Staff Scientist core responsibilities: Algorithm and software development, implementation and testing.
- Design of real time hyperspectral anomaly detection algorithms as well as temperature feature extraction for new LWIR sensor in C++ using Visual .NET and delivered to customer on Linux platform
- Implemented real time PDB search tool to identify structures similar to specified structures (Kinase)
- Implemented and demonstrated Spectral Matched Filter (SMF) algorithm for target detection in a complex target/background scene
- Implemented Atmospheric Correction algorithms in C++ and processed AVIRIS data
Photon Research Associates, Jan. 1996 - April 2012
- Developed algorithms for the IR Sensor Study - Algorithm Development Program that demonstrated advanced detection and identification techniques against measured CBD and CBE data sets
- Participated in development of Cirrus cloud detection and characterization for HYMSMO program
- Developed a mid-course discrimination simulation for SBIRS-Low and performance simulation in order to evaluate the effectiveness of acquisition, tracking, and discrimination in clutter environments
Photon Research Associates, Jan. 1996 - April 2012
- Developed and simulated mid course discrimination algorithms for the Navy Theater Wide COEA (Navy Theater Ballistic Missile Defense)
- Developed Monte Carlo simulations of IR sensor(s) and target scenarios over variations in launch, sensor and solar geometry
- Simulated SBIRS-High, SBIRS-Low and a number of novel staring/scanning sensor designs and demonstrated performance against Technical Intelligence targets, developed code for integrating targets into backgrounds, incorporated a matched filter into high fidelity simulation and control logic for executables that provide parallel processing