British software & AI company · est. 2016

Craftsmen and purveyors of smart systems.

Fizl Ltd is a software house and AI research lab. We deliver full-lifecycle engineering — web, desktop, embedded, networks and hosting — and we build Ariadne, an AI orchestration platform that keeps working after everyone else has gone home. Engineered in Britain, with models trained on our own Grace Blackwell cluster.

2016UK-registered, independently owned
Grace BlackwellIn-house AI training cluster
24/7Ariadne works around the clock

Built on serious infrastructure

Grace Blackwell cluster NVFP4 native inference MSDN Enterprise UK-registered & owned
Our product

Ariadne — one AI orchestration platform for the whole organisation.

Ariadne brings agents, memory, tools and integrations together into one continuous workflow. It holds long-term goals, notices what changes, acts on your systems, and verifies the results — day and night.

  • Goal-driven autonomy. Set goals with success criteria; Ariadne plans and acts, and an independent evaluator confirms the job is genuinely done.
  • One shared workspace. Chat, code, query, notes, whiteboard, kanban and 3D — used by your people and their agents side by side.
  • The MCP context lake. Your organisation's knowledge, exposed over MCP to any compatible agent — inside Ariadne, in your IDE, anywhere.
  • Private by design. Self-host on your own hardware or run it in your cloud. Your data, your models, your rules.
Ariadne — AI orchestration platform

The Ariadne mark — a Conway glider in flight. Emergence, not instruction.

Software development

Full-lifecycle delivery in C#, C++, SQL and TypeScript — web, desktop, embedded and smart-card systems, architected with ArchiMate.

Network infrastructure

Resilient design across L2/L3, BGP, DNS, PKI, Hyper-V, Kubernetes, firewalls and VPNs including MPLS/ExpressRoute.

Hosting & platforms

Windows, Linux and FreeBSD servers, ASP.NET sites, MS SQL and NoSQL — provisioned and maintained.

Consulting & support

Cost savings, efficiency gains and rapid troubleshooting, plus hands-on cloud provisioning across Azure, AWS and Google Cloud.

Beyond the platform

The engineering team behind the AI.

Ariadne is built by a working consultancy. The same team that trains our models can turn your idea into a product, harden your infrastructure, or untangle a difficult IT problem.

Fully equipped with MSDN Enterprise licenses, our own production servers, a secure remote-access solution and specialist contacts across almost every corner of IT.

Foundation models

The models we train in-house.

Trained on our own Grace Blackwell cluster — tuned for speed, intelligence and the ability to act, not just answer.

Diffusion · NVFP4

Kaja-Flash

A native NVFP4 high-speed diffusion foundation model built for real-time, high-throughput generation.

Throughput
700 tg/s
Reference
1 petaFLOP
Best for
Low-latency
Reasoning

Kaja-Base

A high-intelligence foundation model for the heavy reasoning and planning at Ariadne's core.

Throughput
300 tg/s
Reference
1 petaFLOP
Best for
Deep reasoning
Omnimodal

Chikara-Instruct

A fine-tuned NVIDIA omnimodel with text, image, audio and action support for grounded, multi-modal tasks.

Modalities
4
Actions
Native
Best for
Multi-modal

Ariadne itself is model-agnostic — run it on our Kaja models, or point it at any OpenAI-compatible endpoint.

The company

A software house with a research lab attached.

Fizl Ltd is a British software and AI company. When we're not delivering for clients, we're in the lab — building a genuinely adaptive form of AI on our own Grace Blackwell cluster.

We describe what we're building as Level 2–3 synthetic intelligence: systems that don't just recognise inputs, but generate and enact strategies to reach long-term goals — and adjust those goals as the world changes.

Founded and led by Simon Soanes. Read more on the blog, explore our open work on GitHub, or dig into the research behind Ariadne.

🏛️

Est. 2016

UK company no. 10374740, independently owned.

🧪

R&D-first

A working lab exploring self-learning AI beyond machine learning.

🖥️

Own cluster

Models trained on our in-house Grace Blackwell hardware.

🔒

Trust

Private-by-design deployments; your data stays yours.

Get started

Let's build something smart.

A product to deliver, infrastructure to harden, or an AI orchestration platform to run inside your own walls — it starts with a conversation.