Staff Edge Engineer & Edge AI Architect

http://bright.ai

Lead the Stateful AI Hub team: I own architecture and technical strategy for the edge AI execution layer of BrightAI's Physical AI platform (NVIDIA Jetson, Yocto Linux). Scope spans capture, model serving, analysis, and structured reporting for critical infrastructure, including autonomous drone inspection.

• Own the platform's technical direction: requirements, high-level design, ADRs, RFCs, and the interface contracts that let multiple product verticals share one edge platform.

• Architected a C++ inference runtime on the open KServe v2 / Open Inference Protocol standard: multi-backend serving (TensorRT, ONNX Runtime, TFLite, and PyTorch via process-isolated Python workers over shared memory), with crash recovery for unattended field operation.

• Architected the model-pack system: one portable, signed model artifact deployable on edge or cloud, with OTA distribution and provenance from inference back to exact weights.

• Designed a contract-driven analysis engine that converts model output into structured, reportable results, letting new verticals ship as configuration instead of new code.

• Drive hardware-aware optimization, fitting AI models to device compute, memory, and thermal limits.

• Lead cross-program decisions on platform boundaries and edge-to-cloud protocol alignment.

https://www.digi.com/

• Owned the X-ON IoT platform end to end, setting system architecture and technical direction across embedded, edge, and device-to-cloud layers, while aligning the technical roadmap with product strategy to drive platform growth and recurring ARR
• Architected and built end-to-end device-to-cloud systems, spanning core system functionality across devices, edge, and platform layers, including embedded firmware and software, hardware, wireless modules, gateways, and cloud services
• Created and released an open-source XBee C SDK, improving developer experience and enabling consistent integration across device families
• Researched, prototyped, and proposed emerging technologies to product and executive leadership, influencing platform direction and roadmap decisions, including edge AI initiatives
• Partnered closely with Product, Cloud Engineering, RF, Security, Manufacturing, and Supply Chain teams to ensure systems shipped reliably at scale
• Mentored engineers and raised the technical bar through architecture reviews, code reviews, and improved development practices
• Founded and led Digi’s internal AIoT Council, driving cross-functional technical alignment on edge AI strategy, platform integration, and execution readiness across product and engineering teams

• Founded and operate BitWize, an independent embedded systems, IoT, and Edge AI consulting practice supporting startups and product teams from early development through production.
• Provide hands-on consulting across embedded firmware and embedded software, hardware architecture, low-power and solar-powered systems, wireless and RF connectivity, and device-to-cloud IoT integration.
• Consulted on smart infrastructure and energy systems, including a school-zone connected IoT system built on Particle.io and a solar backup charging platform, where I stabilized DSP-based controller firmware to improve system reliability.
• Deliver ongoing and repeat engagements for multiple clients (some confidential), supporting firmware stabilization, system architecture, and product scaling for connected devices and IoT platforms.
• Manage all aspects of the consulting practice, including technical delivery, system architecture, client engagement, project scoping, billing, and finance, operating from a dedicated engineering lab and supporting on-site collaboration as needed.
• Actively expanding BitWize to offer Edge AI and AIoT consulting, focused on deploying ML/AI on embedded and edge devices and integrating intelligence into production IoT systems.

https://www.abb.com/

• Led modernization of industrial automation and controls software across a distributed team, spanning embedded Linux HMI, web-based UI, and backend data acquisition systems used in production controller environments
• Overhauled embedded Linux touchscreen UI stacks, including firmware-level improvements to input handling and rendering performance for industrial controllers
• Re-architected database schemas and Windows-based backend services supporting controller data acquisition and management, improving reliability, performance, and maintainability in production deployments
• Built and integrated production services using Python, SQL, and SNMP, enabling improved device monitoring, control, and data flow across industrial systems
• Led and mentored engineers in a remote / hybrid environment, providing technical direction, code reviews, and hands-on support to unblock execution and deliver delayed projects to production
• Improved development efficiency and software quality across embedded and web teams by introducing stronger engineering practices, tooling, and review standards

• Built and shipped end-to-end LoRaWAN device-to-cloud products, spanning embedded firmware for wireless modules and end devices and embedded Linux (Yocto) software for industrial gateways, forming the core platform that drove Digi International’s acquisition of Haxiot
• Delivered low-power industrial and sensing devices optimized for constrained networks, including proprietary protocol-bridging solutions (e.g., Modbus over LoRaWAN) and a production low-power sensing device with multi-year battery life integrated with the X-ON LoRaWAN platform
• Led technical execution for production readiness, including manufacturing scale-up, regulatory certification, and close collaboration with company leadership to align technical decisions with business priorities in a fast-moving startup environment
• Supported post-acquisition scaling and integration of Haxiot products into Digi’s production portfolio

https://www.quantumiot.com/

• Developed end-to-end IoT systems spanning embedded firmware, sensor integration, cellular connectivity, MQTT, and cloud integration, delivering production solutions across multiple customer verticals (e.g. smart city, asset tracking, etc) in a consulting-driven environment
• Built embedded firmware in C/C++ for connected devices, including vehicle tracking systems publishing OBD, location, and driver behavior data, with support for cellular communication, FOTA/COTA, and reliable cloud ingestion
• Designed and implemented sensor-rich edge devices for industrial and commercial applications, integrating diverse sensors (temperature, humidity, weight, motion, current, cameras) to support monitoring, control, and analytics use cases
• Performed LoRaWAN R&D and prototyping, building private LoRaWAN networks including gateway implementations, sensor nodes, and backend ingestion services to evaluate feasibility for early customer deployments
• Worked closely with company leadership, partners, and customers to translate ambiguous requirements into shippable systems, contributing across hardware bring-up, embedded software, installation, and field testing

• Developed device-side software for digital oilfield automation systems, running on embedded Linux (Raspberry Pi) platforms used for field monitoring and control
• Built Node.js–based applications on edge devices to interface with sensors, automation hardware, and PetroCloud’s existing backend systems
• Integrated edge software with upstream platforms, ensuring reliable data collection, monitoring, and automation in field deployments
• Worked in a fast-moving startup environment, collaborating closely with hardware and platform teams to support rapid iteration and deployment

https://entouchcontrols.com/

• Designed and implemented embedded software for connected energy management products in a fast-moving startup environment, contributing to cloud-connected lighting controls, sensor monitoring devices, and smart thermostats deployed at national retail and restaurant chains
• Developed bare-metal firmware on microcontroller-based platforms used for lighting controllers and remote sensor monitoring devices, integrating sensors and wireless communication to transmit data to gateways and cloud platforms
• Built wireless lighting control and sensor monitoring systems on resource-constrained MCUs, supporting scheduling, events, remote overrides, and sensor-driven automation in commercial environments
• Contributed to embedded Linux–based thermostat platforms, designing HVAC control state machines, implementing PID control algorithms, and developing GTK-based graphical user interfaces for local control and monitoring, alongside cloud-connected features
• Developed web-based tools for device data access, reporting, and remote control, supporting operations, diagnostics, and customer use cases
• Participated across the full product lifecycle, including hardware bring-up, embedded development, testing, manufacturing support, and automated production testing

• Developed embedded C firmware for MSP430-based sensor nodes, supporting data collection from wireless monitoring sensors
• Implemented and modified gateway-side C code to receive, aggregate, and process sensor data from field-deployed devices
• Configured and integrated wireless sensors using TI SimpliciTI–based protocols, supporting network setup and reliable data communication
• Gained hands-on experience with embedded systems, low-power wireless communication, and sensor-to-gateway data flows in an early IoT environment

• Supported manufacturing planning and execution for structural steel products used in highway and infrastructure projects, contributing to on-time delivery of large-scale builds
• Created AutoCAD drawings and programmed CNC machines to cut and prepare steel components to specification
• Coordinated with factory floor teams to translate drawings and schedules into executable production workflows
• Analyzed material requirements, managed inventory needs, and supported procurement to keep production on schedule