Technical Director.

Specialising in A.I and IoT products for smart home, energy, agriculture and automotive sectors.

Intro

I’m a European based technology and product leader with over twenty years experience, specialising in A.I. and IoT products that drive business transformation and value.

My unique "unicorn" skillset combines high-level tech leadership with hands-on hardware and software engineering expertise.

This hybrid approach allows me to:

Quickly validate concepts
I can rapidly design and build prototypes that validate concepts and prove business value, often within the first month.
Fail faster, cheaper
This method significantly reduces time and cost, allowing us to pivot quickly and minimise risk.
Accelerate time-to-market
Once a prototype is proven, I scale the team and bring the product to market in under half the time a traditional product team would normally take.

My background in both software and hardware allows me to lead large teams of multidisciplinary designers, technologists, developers and engineers and having spent the last two decades working at some of the very best creative and technology companies in the world, I’m also able to set the vision for each project.

I’m somebody that enjoys working at C-suite level one minute, then rolling up my sleeves to work hands-on leading the teams back in the studio.

Many people know me from some of the fairly famous connected experiences that I’ve created over the years and I still enjoy coming up with ideas like that from time to time.

Expertise

0-1 Product ideation, strategy, design and build.
Full stack software development.
AI-driven, data-rich digital experiences.
Immersive and web-based platforms.
Cloud & On-prem architecture.
Electronical engineering and hardware design.
Extensive experience building custom A.I models.

Full-Stack & Cloud Architecture

Fullstack web: Javascript, Typescript, React, Node.
Cloud Ecosystems: AWS, GCP, Azure and DevOps/CI/CD practices.
MACH architectures with Next.js, Vercel and headless CMS like Contentful etc.

A.I. & ML

A.I. for Edge devices, with CoreML, TensorflowRT and PyTorch.
GenAI and multi-agentic agents using frameworks like Google ADK and Langchain.
Building and training custom ML models with Tensorflow and PyTorch.
Computer vision using CNN’s like YOLO for object detection and classification.
Quantized models and LLM’s on Edge and mobile devices.

Software

Python, C, C++, Dart, Kotlin.
Fullstack web: Javascript, Typescript, React, Node.
Native mobile development for iOS using Swift and Kotlin for Android.
Hyprid development with Flutter and React Native.

Hardware

CircuitPython, C, C++ specialising in development of ESP32 and Nordic NRF52 bluetooth devices.
Electronic engineering and designing custom PCB’s.
Enclosure design for prototypes and 3D printing.

Bespoke AI-IoT bluetooth device for cow farmers.

Most western cow farms are between 500-1000 cows in size and this kind of scale means farmers can afford all kinds of technology to help them produce higher milk yields. By comparison, the average cow farm in India is around 4 cows and so farmers aren’t able to leverage the same kind of technology as western farmers.

As product designers, this gives us an unbelievable opportunity to design bespoke, low cost tech products for people that the rest of the world has ignored - like this A.I. powered milking device.

Farmers like nice products too! The idea for the enclosure and overall design of this device was to give farmers something beautiful, something that they might not expect from a low cost device and something that would stand-out in market.

I always start projects by rapid protoyping using off-the-shelf electronic components and then combining them on a custom designed PCB. This enables us to make fast interations and test sensor performance quickly, without the costs and timelines attached to creating fully integrated PCB’s.

The best part of designing new products and in my opinion the most important, is that first test with real users. I ask my design teams to do this very early on in every smart product design project and my favourite part is actually iterating and making as many design changes as possible on-site, then testing again. On this occasion that meant updating firmware whilst up to my knees in cow dung!

Designing the next generation OS for electric vehicles.

India has a unique EV market in that most of the vehicles aren’t high end Tesla’s built and designed from the ground up by leading tech manufacturers, but they’re low-end two and three wheeler vehicles built by start-ups. Often these start-ups design and build the vehicle and then struggle when it comes to the display and interface.

For the last 8 months or so, my talented team and I have been designing and prototyping a white-labelled solution for Indian based vehicle OEM’s, consiting of a hardware dash/cluster display and the underlying vehicle OS that interfaces with each OEM’s hardware using the CAN bus.

We’re very excited that months of designing, prototyping and testing have secured our first order of $10m and we are now entering mass production.

The UI design of this project was great but my favourite part of projects like this is always the early testing. In this case, combining the OS, the vehicle and our hardware display/cluster for the first time.

In India, it’s absolutely acceptable to hold everything together with sticky tape if it means a faster and more iterative design process! Maybe this is why they have one of the fastest growth economies in the world at the moment. I’ve definitley never worked so productively as my team and I are doing right now.

Using A.I to help farmers protect their cows.

During my last product testing visit to the farmers based in rural India, I visited one of the cow farms and noticed this cage near to the young cows.  When we asked what it was, we were told it was a leopard trap. It turns out that many Indian farms have a big problem with leopards coming in the day and night to take young cows and farm dogs.  Unfortunately, the leopards were too smart to be trapped, so I decided to think about how we could help the farmers.

CowDefender is an A.I. powered IoT camera that protects cow herds from leopard attacks. Alerts are sent to the farmer whenever a leopard is detected, allowing them to sound the alarm and ward off the leopard.  Detection is done on device in real-time, utilising a custom trained YOLO8 Tensorflow lite model and the latest machine learning edge technology. Featuring an ultra low-light camera, CowDefender can detect leopards even at night, when attacks are most common.

Detect. Record. Alert.

Bringing solar and renewable energy to Indian homes.

My team and I designed the full end-to-end system for Jio’s first foray into the home solar and renewable consumer energy market. Designing, prototyping and piloting software and hardware devices that comprised of solar panels, micro-inverters, powerwall, and home charging products with bi-directional charging together into one cohesive CX.

Crop protection with IoT sensors.

Grape vines are highly sensitive and any changes to soil mositure or other weather conditions, can severely damage vineyards in a short space of time.

We created an IoT crop sensor to warn farmers of potential damage before it happens. Five different sensors measure soil moisture, rain, sun, leaf wetness and atmospheric conditions. Solar panels give enough power to our custom designed PCB and onboard GSM, connecting to the Jio 5G network periodically to log data to the mobile application we designed.

Designing the future of EV charging for India.

I specialise in designing and building connected smart products and I have over twenty years experience designing innovative products for the agricultural, energy and automotive industries.

Creating huge efficiencies in fleet management with smart charging and optimised routing.

I specialise in designing and building connected smart products and I have over twenty years experience designing innovative products for the agricultural, energy and automotive industries.

One of the most iconic billboard adverts of the twenty first century.

From the moment my daughter stopped playing in the garden and started pointing up at the plane that was passing overhead, I knew there was something special about flying in the eyes of a child.

The rest was easy (ish!) I placed an ADSB antenna on the top of a building in Piccadilly and interrupted billboards all across London whenever a British Airways plane flew overhead.

The destination and price for each flight was all used in real-time to create one of the most awarded outdoor campaigns in recent history.

Google Voice Mirror.

Working for Google Brand Studio in London, I worked on a proof of concept voice powered mirror for them to demonstrate the power of Google Assistant and Dialogflow to fashion brands.

The mirror analyzed your look each day and using a custom trained machine learning model, made clothing recommendations based upon your wardrobe, using a specially created DialogFlow A.I. Agent. All through a voice conversation with the user.

Utilising the OpenCV computer vision framework, the camera detected each user and then cropped in on each of the garments they were wearing before sending to the cloud for analysis. This resulted in a much higher accuracy of recognition.

Cloud Vision Glasses.

Back in 2017, before it was really possible to run A.I. or machine learning models on Edge devices, I designed and built glasses for partially sighted or blind people that descibed what was in front of them using Google’s Cloud Vison API. The glasses had a tiny camera inside them that would stream images to the cloud, infer and recognise objects in the images, then return audio back to the user so they could ‘see’.

A.I. driving assistant for Hyundai ‘N’.

A conversational voice app for Hyundai performance brand ‘N’, powered by your driving data. With a difference. Think Kit from Knight Rider but with the character of a boy racer.

If you cornered nicely, ’N’ would squeal with delight. If you were stuck in heavy traffic ’N’ would complain and ask you if you wanted to it to look for a detour and get out of the jam.

From concept to working product, I designed and built the iOS app as well as the cloud server application.

Helping Dogs breathe clean air in cities.

We all know that pollution in cities causes lung cancer and heart disease and most modern cities monitor air quality levels. However, this data usually comes from sensors on the top of buildings, high above the ground where the air has dissipated and is actually far less toxic than on the street, where the more concentrated fumes are harmfully breathed in by people.

With the help of our furry friends, I created a bluetooth dog collar and live data platform that proved just how more harmful and toxic the air is at ground level.

Adidas ‘micoach’ smart boot.

Historically, Nike as a brand has focussed on individual performance and products like Nike+ measure an athletes speed, distance and endurance. As part of a promotional event for the 2014 World Cup, Adidas wanted a similar product, but one that was more aligned to their brand philosophy of teamwork.

I created wifi boots that measured the currency of team play - passing. Using multiple FSR sensors planted in the boot I created a mesh network using ZigBee wireless devices and fed the data into a live data visualisation that was projected on the wall of the cage that the three-on-three street football games were being played in.

‘Champion of the playground’ connected playground products.

A decade before the current trend in fitness and wearable tech, I designed a handful of connected activity toys for Robinsons that rewarded kids for exercising and doing tricks, using accelerometer data inside each toy.

Each product would work hand-in-hand with a virtual online gaming platform we designed. The digital hacki-sacks and skipping ropes outputted a code on a small screen embedded into each product and kids could enter that on the website to work their way up the online leaderboard.

Top scorers were rewarded and champions of the playground found!

Facebook photo frame.

In 2012, I designed and built a 3G picture frame that allowed elderly people and other non social network users, access to all the Facebook family photos that they would normally miss out on.

A user simply dragged an image to a special photo album folder on Facebook and any image they added automatically downloaded to Grandma or Grandpa via 3G.

The device was designed to look like an old analog wireless radio to avoid scaring non-tech-savvy users, with only a simple scroll wheel for moving through the images rather than a touchscreen interface.

Fast forward 10 or so years and Facebook launched an identical product called Portal: https://portal.facebook.com

Much of the work on this site is the property of my clients and some of which is still in development. Please do not share without permission. Thank you.