Editing
I love telling human stories that make audiences lean in: TV ads, product explainers, customer testimonials, paid and organic social.
Motion
I animate brands, guided by their values and creative principles: logos, titles, 3D product demos, apps and websites, in 25+ markets.
Systems
I spot what slows work down or lets errors creep in, build the solution and teach everyone to use it: motion templates, AI tools and shared ways of working.
As a tech company, what3words embraced AI in production as early as 2024. I helped choose the tools, pushed for them and got my team using them. Since then, I’ve used the latest frontier tools every day, in a wide range of ways.
Video assembly tool
One hand-crafted edit becomes a template that can produce dozens of videos, fast and with very little resource. The story is built in blocks, so footage, copy, voice-over and music can be swapped to test variants. Each version can then be localised with translated copy and dubbing. One video I made became 100+ versions across multiple markets.
Tools: Creatomate for the template, Claude for copy and translation, ElevenLabs for voice-over, Synthesia for dubbing, and custom software for the render queue.
Generative video and image
Photoreal video and images, anchored in the real world. Creating AI presenters (avatars), putting a presenter into a new environment, creating drone shots from stills, changing the season and colours in footage, reframing between horizontal and vertical, and extending and upscaling to fit the delivery format.
Tools: Nano Banana in Google Flow, Seedance 2.5 in Runway, HeyGen and Synthesia for AI avatars, Topaz for upscaling, Claude with custom skills for writing complex prompts.
AI Audio
Synthetic voice-over and dubbing, clean-up for badly recorded sound, and transcription for subtitles.
Tools: ElevenLabs for voice-over, dubbing and clean-up, plus Synthesia for dubbing, Adobe Podcast for clean-up and Whisper for transcription.
Long before AI, I was automating the repetitive bits.
Why I built it
Brands are built on repetition: the same elements, across all videos. But each job needs them with new copy, colours, aspect ratios or languages. Rebuilding them by hand is slow and easy to get wrong.
What I did
Since 2018, I’ve built data-driven After Effects templates and Motion Graphics templates (MOGRTs) for Premiere Pro. They lock the design, timing and brand rules, and adapt to whatever the job needs.
Result
At Technicolor, one template produced 25 localised versions in 2 minutes, not 30. At what3words, 20-minute builds became a drag and drop. And because the template gets the brand right every time, four editors on two continents work as one team.
Where we started
At what3words, half my team worked in Mongolia, 8 hours ahead of London. Editors worked in silos, and nobody could see how anyone else got to their final cut. I wanted one team that knew each other’s work and kept growing together.
What I did
I made sure the editors in Mongolia were as much part of the team as those in London. Every month, we came together to celebrate the best work and share what inspired us. In weekly meetings and one-to-ones, we talked about where each person wanted to grow. I shared the technical know-how to get them there, and got them teaching each other too. And I took the friction out of their day. One shared drive, with the same structure for every project, meant anyone could pick up anyone’s work. Handover shrank from hours of preparation to one templated Slack post, so everyone could focus on the craft.
Result
Two mid-weight editors grew into senior roles, and I supported one of them as he managed a junior for the first time. One of our editors, who started out cutting video only, went on to create complex motion graphics. And a quarter of our projects were handed over or made together, with not one missed deadline.