The Agent Reasons, the Engine Remembers: Designing an Agentic Workflow EngineJul 7, 2026·21 min read·25
Prompt and Model VersioningOr: The day the diff was empty and the behaviour wasn'tJul 14, 2026·15 min read
The Intern with a Refund Button: Designing a Customer Support Agent at ScaleJun 30, 2026·14 min read·12
The Retriever Is Your Security Perimeter: Designing a Multi Tenant RAG Knowledge BaseThe moment an LLM starts narrating retrieval results, your retriever becomes your access control layer. Filter permissions anywhere above that and you have built a very sophisticated data leak with excellent citations.Jun 25, 2026·18 min read·18
Faster Than the Fingers: Designing a Real Time Coding Assistant*When the model sits in the loop of a typing user, the job stops being "make the model smarter" and becomes "make a model that is slower than the user feels faster than the user." Good luck.*Jun 23, 2026·18 min read·14
Designing AI Search: The Answer Layer and the Iceberg Beneath ItAn LLM synthesis layer did not replace the search iceberg. It made every layer below it works harder: fast enough to feed the model, fresh enough to stay true, and structured enough to cite.Jun 18, 2026·20 min read·13
More Agents, More ProblemsWhy multiagent systems are just distributed systems wearing a trench coat, and how to architect them so they survive past the demo.Jun 16, 2026·14 min read·32
Your Agent Has a Spending ProblemCost is an architectural property, not a billing surprise. Treat your agent as a metered process, or a few sessions will eat a month of budget in a weekJun 11, 2026·15 min read·13
In Probabilistic Systems, You Watch the Shape of SuccessAll green. All wrong..Jun 9, 2026·18 min read·42
