Blog

Notes from the field.

In-depth articles on building AI systems: agents, platforms, sovereign compute, migrations and applied research.

11 min read

Decision models: the missing layer between LLMs and code

Jev and its open alternatives return typed decisions with probabilities instead of text. What they are, what the first evidence shows, and where a decision layer fits in agent architecture.

Decision ModelsJevAI AgentsArchitectureCalibration Read →

11 min read

Fact-retrieval vs extrapolating AI agents: the difference, and how to build both

Some questions have an answer in your data; others need an estimate of what isn't known yet. Why retrieval agents alone fail at forecasting, what-if and risk questions, and how to build agents that do both.

AI AgentsRAGForecastingArchitectureEvaluation Read →

10 min read

97% of frontier-LLM quality at up to 1,000x lower cost

Not every AI task needs a model that talks. Our hybrid ModernBERT pipeline processes tens of millions of texts in 30 minutes for under $10 a day, at 97% of frontier-model quality.

ModernBERTLLMDistillationJevMLOpsCost Optimisation Read →

9 min read

Horizontal vs vertical AI agents is the wrong question

Scope is only one of five design choices for enterprise agents. How to decide ownership, execution authority, mechanism and architecture, with the governance that makes them safe.

AI AgentsArchitectureGovernanceEnterprise AIStrategy Read →

11 min read

How to run coding agents securely: sandboxes, microVMs and defaults

Coding agents run shell commands, install packages and reach networks. How isolation tiers compare, what Copilot, Codex, Gemini, Grok Build, OpenClaw, Azure, E2B and Daytona offer, and a setup that holds up.

AI AgentsSecuritySandboxingCoding AgentsAzureE2BDaytona Read →

8 min read

The AI-native SDLC: how to adopt it without losing control

Coding agents make the build phase fast. Planning, review and deployment are now the bottleneck. A practical guide to redesigning your software lifecycle, with governance built in.

AI AgentsSDLCGovernanceClaude CodeEngineering Read →

Next step

Let's talk about your AI system.

A free 30-minute call with an Engagement Lead or AI Architect. You'll leave with a clearer view of options, risks and cost, whether or not we work together. Your case doesn't need to fit any box on this site; just tell us what you're facing.