Lead Engineer, Drevise
Full-stack applications
Cloud architecture
Teams & products that just work
NestJS · AWS
I build teams and products that just work.
I'm Mo Mia — CTO of CapSquid and lead engineer at Drevise, based in Amsterdam. I've led engineering teams inside a publicly traded manufacturer and built AI products from scratch. I help companies make the technical decisions that are expensive to get wrong, then build the teams that carry them out.
For founders, CTOs and engineering teams who need an experienced outside view on what to build, who should build it, and what it will cost to run.
The team decides whether the product works. I help you hire the right engineers, set up how they work together, and get to a team that ships reliably — whether you're making your first hires or fixing one that has stalled.
Part-time technical leadership before you have it full-time: roadmap, architecture and build-versus-buy calls, and running engineering until a permanent leader is in place.
Where language models and agents genuinely help your product and your team, and where they don't. Evaluation, cost, data handling, and agentic workflows engineers keep using.
A structured look at your system design, infrastructure and cloud bill — serverless or containers, Kubernetes, workflow engines — with a written recommendation you can act on.
What bigger customers ask before they sign: SSO, on-premise deployment, audit trails and compliance groundwork such as SOC 2 — sequenced so it doesn't stall the roadmap.
The team decides whether the product works. I help you hire the right engineers, set up how they work together, and get to a team that ships reliably — whether you're making your first hires or fixing one that has stalled.
Part-time technical leadership before you have it full-time: roadmap, architecture and build-versus-buy calls, and running engineering until a permanent leader is in place.
Where language models and agents genuinely help your product and your team, and where they don't. Evaluation, cost, data handling, and agentic workflows engineers keep using.
A structured look at your system design, infrastructure and cloud bill — serverless or containers, Kubernetes, workflow engines — with a written recommendation you can act on.
What bigger customers ask before they sign: SSO, on-premise deployment, audit trails and compliance groundwork such as SOC 2 — sequenced so it doesn't stall the roadmap.
Technical leadership at two growing companies and a publicly traded manufacturer — from setting the architecture to running the teams that build it.
A European platform that turns commits, issues and payroll data into audit-ready software capitalisation and R&D tax-credit reporting — no timesheets.
An AI platform that reads technical drawings and specifications, automating the path from engineering review to series production for manufacturers and their suppliers.
Led two engineering teams at a publicly traded digital manufacturer, stepping in as their engineering manager when none was in place.
A European platform that turns commits, issues and payroll data into audit-ready software capitalisation and R&D tax-credit reporting — no timesheets.
capsquid.com ↗An AI platform that reads technical drawings and specifications, automating the path from engineering review to series production for manufacturers and their suppliers.
drevise.com ↗Led two engineering teams at a publicly traded digital manufacturer (NYSE: PRLB), stepping in as their engineering manager when none was in place.
protolabs.com ↗Models, agents and apps I've built or contributed to under the Otron name — where I test ideas before recommending them to anyone else.
A fine-tuned ModernBERT-large classifier that answers several typed questions about a text — yes/no, pick-one, or level — in a single local forward pass, in roughly 16–25 ms on Apple silicon.
An autonomous development agent across GitHub, Slack and Linear: it answers webhooks, manages issues, opens pull requests and remembers what it has done.
A private menu-bar utility: select any region of the screen and its text lands on your clipboard. Apple Vision on-device, no cloud.
A menu-bar app that turns any web page into a live desktop wallpaper, and tints the menu bar to match it.
A 3D machine-shop game in the browser: run timed CNC operations and fill orders. Built by a friend; I designed and built a new game mode.
A fine-tuned ModernBERT-large classifier that answers several typed questions about a text — yes/no, pick-one, or level — in a single local forward pass, in roughly 16–25 ms on Apple silicon.
An autonomous development agent across GitHub, Slack and Linear: it answers webhooks, manages issues, opens pull requests and remembers what it has done.
A private menu-bar utility: select any region of the screen and its text lands on your clipboard. Apple Vision on-device, no cloud.
A menu-bar app that turns any web page into a live desktop wallpaper, and tints the menu bar to match it.
A 3D machine-shop game in the browser: run timed CNC operations and fill orders. Built by a friend; I designed and built a new game mode.
I trained in computer engineering and have led engineering from early-stage startups to a publicly traded manufacturer — product, infrastructure and, more recently, AI systems including models I've trained myself.
Originally from South Africa, I've lived and worked in Amsterdam since 2021. I run engineering day to day, so the advice I give comes from systems I'm responsible for now, not ones I read about.
Send a few lines about your company, the problem and your timeline. I'll reply within two working days.