AI
Shipping LLM features that survive contact with real users: evals, agents, retrieval and restraint.
AI cost observability: tokens are a line item now
LLM features live or die on unit economics. Build per-feature token budgets, cost-per-successful-task dashboards, cache tracking and frontier charts execs actually read.
Agent UX: designing for software that acts on your behalf
When software acts instead of suggests, the interface becomes a control system: permission scopes, visible progress, real undo, and calibrated trust.
Showing the swarm: status UX for multi-agent systems
Multi-agent features bury users in log firehoses or leave them staring at a spinner. Designing step summaries, cost displays, interruption and failure narration.
Latency budgets for LLM features
An LLM feature that answers correctly in eleven seconds is broken. How we set per-surface latency budgets, route models for speed, and kill what can't make the time.
Fine-tuning vs RAG vs prompt: a decision guide
Prompt, retrieve or fine-tune? The decision tree we run with clients — real cost curves, data needs, iteration speed and maintenance burden.
Shipping LLM features: a production checklist
The pre-launch checklist we run before any LLM feature ships: evals, traces, cost ceilings, fallbacks, feedback loops and the metrics that catch quality drift.
Evals are the new unit tests: a practical guide for product teams
Golden sets, rubric graders, regression gates and CI plumbing: how product teams build LLM evaluations that catch real regressions without a research team.
Feedback loops users will actually feed
Thumbs-down rates near zero tell you nothing. Designing AI feedback signals that collect usable data, close the loop visibly, and track volume as a health metric.
When the model fails: fallback UX for LLM features
LLMs fail — slowly, confidently and at the worst times. How to design degradation tiers, honest error states and timeouts that keep users' trust intact.
Shipping LLM features: what production taught us
The lessons that only arrive after launch: latency budgets in the wild, streaming UX edge cases, cost surprises, fallbacks, and expectations users hold.
AI and brand voice: keeping the robot on-script
Models drift toward LinkedIn-speak. How we write voice specs machines can follow — exemplars, register maps, persona rules — and test them with real readers.
Agent UX: designing for delegation, not magic
Agentic features fail when they promise magic. The interface patterns we use instead: scope setting, visible progress, checkpoints, undo and earned autonomy.