Field notes
The AI Ascend blog.
Deep dives, applied tutorials, and the AI news that matters.
Jun 02, 2026
Production Agent Engineering: The Pro-Code Playbook
Fifty-seven percent of organizations now run AI agents in production, yet quality remains the top deployment barrier. This is the field guide for the engineers shipping those systems: frameworks compared, patterns catalogued, failure modes documented, and costs quantified.
27 min read
Jun 02, 2026
LangGraph: Stateful Agent Orchestration from First Principles
Most agent frameworks treat state as an afterthought, tacking memory onto a chain of LLM calls. LangGraph inverts the design: the graph is the state machine, every node transition is checkpointed, and cycles are first-class citizens. That inversion is why Uber, LinkedIn, and Klarna run their production agents on it.
24 min read
Jun 02, 2026
Microsoft Azure AI Foundry: The Enterprise AI Development Platform
Microsoft has renamed its AI development platform three times in three years, from Azure AI Studio to Azure AI Foundry to Microsoft Foundry. Behind the branding churn is a genuinely ambitious consolidation: 1,900+ models, a managed agent runtime, built-in evaluation, and a resource hierarchy designed for enterprise governance. This is the A-to-Z guide.
24 min read
Jun 02, 2026
The Anthropic Platform Stack: An Architect's Guide to Building on Claude
Most teams still treat Anthropic as 'the Claude API.' That framing misses the platform that has grown around it: managed agents, a universal integration protocol adopted by every major AI vendor, a governance layer with 28 security integrations, and deployment patterns that are replacing traditional application architectures. This is the reference map for architects building on that stack.
27 min read
Jun 02, 2026
The Agentic Runtime: Why the Orchestration Layer Is Becoming More Valuable Than the Model
A frontier model can write a function. An agentic runtime can read a codebase, plan a migration, edit forty files, run the tests, fix what broke, and open a pull request. The difference is not intelligence; it is the system that turns intelligence into sustained action.
27 min read
Jun 01, 2026
The Hidden Cost of AI: Why Inference Is the New Cloud Bill
Training a frontier model is a one-time expense measured in millions. Serving it to users is a perpetual expense measured in billions of tokens per day. For most organisations, inference is already the larger number, and the gap is widening.
23 min read
Jun 01, 2026
From Encoder-Decoder to GPT-2: How the Transformer Learned to Just Decode
The 2017 Transformer was built to translate, with a symmetric encoder and decoder. Two years later GPT-2 threw half of it away and got better at language. The reason lives inside one matrix multiplication.
22 min read
May 31, 2026
What the bake-off taught us: classical ML is not dead, it is just under-attended
We pitted twelve sklearn algorithms head-to-head on a tabular dataset. The winner was not the most expensive one. It was not the most modern one. It was the one whose assumptions matched the data.
3 min read
May 31, 2026
Prompt caching is not just a cost optimisation - it changes what apps you can build
Most teams treat prompt caching as a billing hack. The deeper consequence is that you can now ship product patterns - persistent agents, large-context tools, system-prompt-heavy UX - that were latency-prohibitive eighteen months ago.
4 min read