Insights & Updates
Explore insights from our engineering team on software development, AI, cloud architecture, and industry best practices.
Master Claude API integration for enterprise applications. Learn authentication, streaming responses, tool use, vision capabilities, and best practices for building reliable AI-powered features with Anthropic's most advanced models.
Jishu Labs
CTO
January 20, 2026
13 min read
Featured Article
10 articles tagged “engineering practice”
Clear tag filterTechnical debt is the future cost of a shortcut taken now. The metaphor is precise about one thing most teams get wrong: debt is only a problem when you stop servicing the interest.
When 41% of code is machine-written, the skills that compound are not the ones most training still teaches. What to prioritise, what quietly lost value, and how to build a path that survives the next model release.
Transcription is solved; the value is not in the transcript. A practical pipeline for converting recorded conversations into durable, retrievable decisions instead of another archive nobody reads.
Onboarding used to mean finding the person who wrote it. In a codebase where much of the code was generated, that person may not exist. A practical approach to ramping engineers into code with no author.
With a large share of new code now AI-generated, review is the bottleneck rather than authoring. What machines review well, what they review badly, and how to split the job without losing the part that mattered.
A story point is a relative measure of effort, not a unit of time. Almost every problem teams have with them comes from quietly converting them back into hours.
A PRD states what should be built and why, precisely enough that engineers can build it without guessing. In a world where implementation is cheap, the specification became the constraint.
A code smell is a surface indication that usually corresponds to a deeper problem. The word 'usually' is the whole point, and it is the part that gets dropped when smells become lint rules.
Four measures of software delivery performance: deployment frequency, lead time, change failure rate and time to restore. They work because they are hard to game together - and they break the moment one becomes a target.
When implementation is cheap and fast, the specification becomes the bottleneck and the durable artefact. What changes when the spec, not the code, is the thing you maintain.
Get the latest insights on software development, AI, and technology trends delivered to your inbox.
Get consistent results. Collaborate in real-time.
Build Intelligent Apps. Work with Jishu Labs.