Hi, I'm
Ben Remnant
Senior Development Engineer
I turn real-world constraints into reliable digital systems, from soldering irons to cloud consoles and everything in between.
About Me
My work spans the entire technology stack, from custom PCB design and oscilloscopes to REST APIs and production deployments. I'm just as comfortable writing embedded firmware as I am building IoT networking infrastructure, architecting a web application, or sitting in a product design meeting with a client.
I'm a strong advocate for using AI tools to accelerate development. I work with large language models, inference hardware, and custom automation every day.
I believe software development will be largely commoditised in the next few years as AI improves. When everyone can generate code, the real value is in knowing what to build and how to structure it. I can architect a system end to end: choose the right MCU for a power budget, design a protocol layer that handles unreliable connections gracefully, structure a backend that scales, and make the tradeoffs that only come from understanding every layer of the stack.
Vintage
30 years old (June, 1995)
Location
Christchurch, New Zealand
Interests
Aesthetic UX, AI Frontier, Automation
Projects
A few things I've built, from embedded hardware to web apps.
PulseLink
Real-time IoT telemetry dashboard. ESP32 sensor nodes push data over MQTT to a Node.js backend with live WebSocket charts and threshold alerts.
View Details →Frostbyte
Edge AI vision system for cold-chain monitoring. Runs object detection on a Jetson Nano to flag open doors, temperature anomalies, and stock levels.
View Details →SignalBox
Custom LoRaWAN gateway and sensor network for remote environmental monitoring. Battery-powered nodes report soil moisture, temperature, and rainfall.
View Details →Homebase
Self-hosted smart home hub that ties together BLE sensors, relays, and cameras with a local-first web dashboard. No cloud dependency.
View Details →Batchline
Firmware build and OTA update manager for embedded device fleets. Tracks versions, stages rollouts, and flags failures across hardware revisions.
View Details →Gridware
Web-based PLC control panel that reads Modbus registers over RS485 and displays live process data with configurable alarm thresholds.
View Details →Skills & Tools
A broad toolkit built across years of hands-on engineering.
Hardware & Firmware
I design and develop custom PCB hardware and embedded firmware for a range of microcontrollers. From bare-metal C on STM32s to rapid prototyping with ESP32 and Arduino, I build reliable systems that operate at the edge.
Networking & Protocols
I work across a wide range of communication protocols, from short-range serial buses and Bluetooth through to cellular, satellite and LPWAN technologies for remote IoT deployments.
Languages
I'm comfortable across the stack, writing firmware in C/C++, building backends in Node.js, scripts in Python and leveraging GPU compute with CUDA. I pick the right language for the job and move between them fluently.
Cloud & Infrastructure
I deploy and manage production systems on AWS, handle domain and DNS configuration, server hardening and ensure data security with proper encryption practices from device to cloud.
Tools & Databases
My daily toolkit includes modern editors, version control, and containerisation. I choose databases based on the workload, from lightweight embedded SQLite through to managed cloud services like DynamoDB.
Artificial Intelligence
Daily Assistant
ChatGPT 5.2
Architecture & Coding
Claude Opus 4.6
Daily-Driven Coding
Claude Sonnet 4.5, GPT 5.3 Codex
Small Tasks
Gemini 3 Flash, GLM-5, Kimi 2.5
AI Workflow
AI is a core part of my workflow. I use Cursor as my primary IDE, with experience in Claude Code and Codex. I've also built scripts and automation pipelines to streamline how I interact with models, from prompt chaining to automated code review. For local inference I run models on my RTX 3080 & 3090 GPUs, and scale to the cloud via OpenRouter and RunPod when heavier compute is needed.
Training & Adoption
I've trained our junior engineers to use AI tools effectively in their day-to-day work, from prompt engineering basics to integrating LLMs into their development workflow. I've also worked with marketing and management teams to identify where AI can save them real time: drafting content, summarising reports, automating repetitive tasks, and building lightweight internal tools. The goal is always practical adoption, not hype.
Edge AI
I build and deploy AI systems that run entirely on edge hardware, no cloud required. I work with platforms like the NVIDIA Jetson Nano, Hailo-8 AI Accelerator, and Raspberry Pi to run real-time inference for computer vision, voice pipelines, and local LLMs. Keeping models on-device means lower latency, better privacy and improved security.
Image Generation
I've supported our marketing team by generating high-quality visuals with AI image generation models. I began with local models like SDXL and Flux, moved to gpt-image-1 for stronger prompt adherence, and finally settled on the SOTA Nano Banana Pro for product images, enabling rapid creative iteration without traditional design bottlenecks.
User Experience
I care about how things feel to use, not just how they work under the hood.
I've spent enough time watching people struggle with bad software to know that a technically correct solution isn't always a good one. The best interface is one the user never has to think about. That belief shapes every product decision I make, from button placement to system architecture.
Functional First
Every element should earn its place. If it doesn't help the user get something done, it goes. Interfaces exist to serve people, not to look impressive in a portfolio.
Clean & Simple
Strip out the noise. A clean layout with clear hierarchy lets users focus on the task at hand. Complexity in the backend should never leak into the frontend.
Self-Explanatory
If someone needs a manual to use an app, the design failed. Labels, layout, and flow should guide users naturally. Good UX is invisible.
Space to Breathe
Cluttered screens overwhelm people. White space isn't wasted space. Giving elements room to breathe makes everything easier to read, find, and use.
Built for Everyone
Good UX means accessible UX. Proper contrast, keyboard navigation, semantic markup. These aren't extras, they're the baseline.
Fast & Responsive
A beautiful interface that takes five seconds to load is a bad interface. Performance is a UX feature. So is working properly on a phone.
Work Experience
Where I've been and what I've built along the way.
Senior Web3 Developer
Blackout GamesPLC Software Engineer
Advanced Motor ControlFreelance Web Developer
Self-EmployedBSc, Computer Science
Victoria University of WellingtonI built a deep understanding of how computers actually work, from data structures and operating systems to networking and algorithms. This was also my first real exposure to machine learning: I built a classifier for the Iris dataset and was immediately hooked. Reading Ray Kurzweil's "The Singularity Is Near" around the same time sealed the deal. I've been obsessed with AI ever since.
Where It Started
Early Projects & TinkeringI got hooked on electronics as a teenager. I taught myself Arduino, basic circuits, and enough C to be dangerous. I built many projects, including an autonomous RC car and a solar panel that tracked the sun across the sky. Most of it held together with duct tape, hot glue and cardboard. The tinkering never stopped. I still spend my free time building personal projects, learning new tools, and pulling apart anything I can get my hands on.
Get In Touch
Have a project in mind or a job opportunity? Drop me a message.