Platforms and software we have shipped
Web applications, mobile apps, and full SaaS platforms. Custom work for clients, alongside the products we build and run ourselves.
The difference custom software makes
Good software should pay for itself. The applications and automations we build for clients are designed around a business outcome, and these are the outcomes they delivered.
How we build
We build software that fits your existing processes rather than forcing you to change them. Every application is maintainable and scalable, and built around how your team works, so it keeps performing as you grow.
You own what we build. The code, the infrastructure it runs on, and the decisions behind both are handed over rather than held hostage.
- User centered developmentWe start by understanding your workflow, then build software that fits the way your team already works instead of forcing them to change.
- Quality assuranceTesting, security review, and performance work before it ships, so the software holds up once other people depend on it.
- Ongoing partnershipMonitoring, updates, and new features after launch. Software that nobody maintains stops being an asset fairly quickly.
Our development stack
What we reach for by default. If your team already runs something else, we work in that instead.
- React and Next.js
- TypeScript
- Tailwind CSS
- Progressive web apps
- Node.js and Python
- PostgreSQL and Redis
- REST and GraphQL APIs
- Microservices
- Docker and Kubernetes
- Carpathian Cloud
- CI/CD pipelines
- 99.5% uptime SLA
- OWASP standards
- Automated testing
- Penetration testing
- Vulnerability scanning
Evaluate My Title
An AI content optimization platform built for marketing teams. It analyses content performance across digital channels and suggests changes, so a team can see what is working before they commit more budget to it.
Built for clients, run by them
Most of what we deliver is not ours to show. Client platforms run under their own name on their own infrastructure, which is the point.
What we can show is the shape of the work: the outcome it was built for, the stack it runs on, and what happened after it shipped.
Open source projects
We build and maintain open source software the same way we build for clients. These projects are free to use and available on GitHub. More on the open source page.
Veritate
A hand-coded byte-level language model inference engine written in C and architecture-specific assembly. Every kernel is custom, and it ships with a dashboard for watching the model work while it generates.
View the projectFree Signature Co
Professional email signatures with no signup, no login, and no payment. Pick a template, fill in your details, and paste into Gmail or Outlook. The whole tool runs in your browser.
View the projectSynevyr
An open source harness for distributing AI workloads. It connects remote, local, and region-separated models over one secure link, so a workload can be sent to whichever model should handle it. Available under the MIT license.
View the projectHave something you need built?
Tell us the outcome you are after rather than the feature list. We will tell you what it takes to get there and what it will cost.