The eye behind the interface
What photography can teach us about attention, restraint, and designing software that feels intentional.
3 min read
Between the lens
and the logic.
What photography can teach us about attention, restraint, and designing software that feels intentional.
A little more
perspective.
How we frame a photograph. How we design an interface. How we build a system that earns trust. Different disciplines, connected by attention to detail.
Explore the collectionA practical way to separate missing evidence from weak answers—and make RAG improvements measurable.
Timeouts happen. Retries happen. Idempotency is how an API remembers what the user actually intended.
Designing movement with a purpose: continuity, feedback, and a genuinely considered reduced-motion experience.
Follow a curiosity. Stay for the details.
What photography can teach us about attention, restraint, and designing software that feels intentional.
A practical way to separate missing evidence from weak answers—and make RAG improvements measurable.
Why splitting documents well requires preserving meaning, not finding a universal token count.
Timeouts happen. Retries happen. Idempotency is how an API remembers what the user actually intended.
Moving work into a queue is the beginning. Durable state, safe retries, and honest progress make it dependable.
Fast is only useful when the data is still appropriate. Think freshness, ownership, and invalidation before hit rates.
Follow a request across services with traces, and turn a vague report of slowness into a useful question.
A useful automation knows what it may do, when it must ask, and how to stop with an honest result.
Designing movement with a purpose: continuity, feedback, and a genuinely considered reduced-motion experience.
Move beyond happy-path examples and protect the behaviour another person or service is actually relying on.
A practical look at trust boundaries, prompt injection, and limiting what an LLM application can access or change.
Learn how to build robust RAG systems with vector databases, embedding optimization, and hybrid search approaches for maximum accuracy.
Building efficient CV pipelines for video annotation, object detection, and frame-level analysis using OpenCV and modern frameworks.
Essential patterns for building robust FastAPI backends with authentication, validation, logging, and error handling.