» Tag
ai-engineering
15 postsWhen an LLM's Training Data Expires: Fixing a Car Pricing Engine
An AI car-pricing engine broke because its LLM recalled stale exchange rates, revealing risks of mandatory retrieval and cache provenance in LLM apps.
Engineering Production Agentic Systems, Part 2: Guardrails
Part 2 of a field manual on production agentic systems: tool surface design, authorization scopes, and audit-trail engineering explained.
Beyond Context Engineering: A Discipline for Reliable LLMs
A position paper argues LLM reliability requires channel engineering, not just context engineering, and introduces the Socium collaboration model.
CommitBrief — AI code reviews, right in your terminal
A provider-agnostic, local-first CLI that reviews your staged changes, a historic range, or a whole GitHub pull request. Zero telemetry, no server. Free and open source.
commitbrief.comHow Airbnb Cut LLM Evaluation Time From Weeks to a Day
Airbnb's four-layer LLM evaluation stack uses deterministic caching and micro LoRA adapters to cut iteration time from weeks to a single day.
Stop Editing Prompts, Build a Context Compiler Instead
Instead of hand-editing prompt strings, treat context assembly as a compiler: versioned templates, testable rules, and full traces of what shipped.
Agentic Workflows Should Mature Into Deterministic Code
New research shows agentic workflows should mature into deterministic code as patterns repeat, using LangGraph as the visible seam.
What 95 Generative AI Job Postings Reveal About Hiring Today
An analysis of 95 US Generative AI job postings shows RAG and agents converging, Python-LLM-RAG as the top skill combo, real salary data, and where hiring actually happens.
Multimodal Models Fail at Sampling, Not Understanding
Multimodal models aren't bottlenecked by capability but by sampling defaults—frame rate, chunking, cropping—that silently limit what they perceive.
Cutting Claude API Costs 84% by Separating Vision and Reasoning
An OCR pipeline routed page images to Haiku and reasoning to Sonnet instead of Opus, cutting Claude API costs by 84%. Includes the model-routing math and a real JSON parsing bug.
Real-Time AI Telemetry Streaming with Genkit and Angular Signals
How to stream multi-step Firebase Genkit AI pipeline progress straight into Angular Signals without database writes, replacing generic loading spinners with live status UI.