Intelligent software built to be useful.

We design and develop practical digital products, from web applications and games to automation and AI-powered tools.

About

Software that earns its place.

We build practical software: web and mobile applications, games, automation, backend services and AI-enabled products. We work end to end — from the first sketch and data model through to a deployed, maintained system that people actually use.

We are small and hands-on by design. You talk to the people writing the code, decisions get made quickly, and the work stays understandable long after launch. Not every problem needs a new platform; sometimes the right answer is a small tool, a tidy API, or a script that quietly removes an hour of manual work each day.

Services

What we build

A full stack of capability under one roof, so a product does not have to be handed between teams to get finished.

Web & app development

Responsive web applications and mobile apps built on modern, well-supported stacks — with real attention to performance, state and offline behaviour.

Game development

Gameplay systems, tooling and full titles — prototypes that prove the loop first, then production code, content pipelines and builds for the target platforms.

AI integration & intelligent features

Language models, classification, search and assistance wired into products where they measurably help — with evaluation, fallbacks and sensible cost control.

Local & cloud AI applications

Deployments that fit the data: models running on your own hardware when privacy or latency demands it, hosted APIs when scale and speed matter more.

Prototyping & MVP development

A working version in weeks, not quarters — scoped to answer the question you actually have, and structured so the good parts survive into v1.

Backend & API development

Services, data models and integrations designed to be readable and observable — clear contracts, migrations you can trust, and logs that answer questions.

Automation & custom tooling

Internal tools, pipelines and integrations that take repetitive work off people's desks — built to be maintained by your team, not just by us.

How we work

Pragmatic, maintainable, shipped

Pragmatic by default

We choose boring, proven technology unless there is a concrete reason not to. The interesting part of a project should be the problem, not the stack.

Built to be maintained

Readable code, documented decisions and tests where they pay for themselves — so the next developer, including yours, can keep moving.

Secure and privacy-conscious

Least privilege, sane defaults, encrypted secrets and data kept where it belongs. Local-first or self-hosted whenever that is the better fit.

Focused on shipping

Short cycles with something working at the end of each one. Feedback from real use beats another round of speculation.

Selected work

Recent projects

A sample of the kinds of systems we design, build and keep running.

Web platform

Card design & publishing platform

A full-stack application for creating, organising and exporting card-based content, with a Python API, a structured asset pipeline and a companion mobile client.

  • FastAPI
  • PostgreSQL
  • Kotlin

AI system

Private knowledge & memory service

A self-hosted retrieval and memory layer that keeps project knowledge searchable across tools, with scoped access, provenance tracking and no data leaving the network.

  • Retrieval
  • Local models
  • MCP

Tooling

Media generation pipeline

A browser front end over a queued generation backend — job orchestration, previews and asset management for producing visual content at volume.

  • React
  • Queues
  • Model APIs

Games

Tabletop & digital game tools

Generators, content editors and companion apps for game groups and small studios, built around fast iteration on rules, content and layout.

  • Game logic
  • Content tooling
  • Export

Artificial intelligence

AI where it does real work.

We treat models as components, not as a product category. That means picking the smallest thing that solves the problem, measuring whether it actually does, and being honest when a plain database query would do the job better.

  • LLMs inside applicationsDrafting, extraction, summarising and structured output, integrated into existing flows.
  • AI-assisted workflows & automationHuman-in-the-loop steps that cut manual handling without removing oversight.
  • Local & private deploymentsSelf-hosted models for sensitive data, regulated environments or offline use.
  • Retrieval, memory & knowledge systemsSearch, embeddings and context stores that keep answers grounded in your own material.
  • AI APIs & model integrationsProvider integrations with caching, rate limiting, retries and predictable cost.
  • Intelligent internal toolsAssistants and dashboards for support, operations and research teams.

Contact

Have something you want built?

Tell us what the product needs to do and where it is stuck. We will tell you honestly whether we are the right people for it — and what it would take.