Blog

AI agents, RAG, and integrations — notes from production.

How a Shelf-Monitoring Computer Vision Pipeline Is Architected

Shelf-monitoring computer vision architecture: on-device detection, a portable runtime, pub-sub transport, and confidence-based routing to a review queue.

Edge vs Cloud Inference for In-Store Computer Vision

Edge vs cloud computer vision for retail: store links can't upload shelf photos fast enough, so detection runs on-device and only the result leaves the store.

Giving Merchandising Control of Search Relevance Without Engineering Tickets

Merchandising control of search relevance: boost, bury, pin and filter rules stored as data the team edits directly, with owners, expiries, and an audit log.

Postgres + pgvector vs a Dedicated Vector Database for Product Search

pgvector vs a dedicated vector database: at catalog scale, keeping embeddings in Postgres wins on consistency and ops; the crossover is millions of vectors.

Hybrid Search: Combining Keyword and Semantic Retrieval in One Ranking

Hybrid search runs keyword and semantic retrieval together and fuses the results with reciprocal rank fusion, so exact SKUs and descriptive queries both work.

A Checklist for Choosing a Site Search Vendor for a Large Catalog

A site search vendor checklist: relevance tested on your own catalog, index freshness against a fast feed, merchandising control, and pricing at scale.

Build vs Buy: Managed Search Platforms vs a Custom Semantic Search Stack

Build vs buy for semantic search: a managed platform, a full in-house build, or a custom stack on a managed embedding API and a vector extension.

What to Log From an AI Support Agent — and Why the Audit Trail Matters

An AI support agent audit log serves three jobs: defending an action later, debugging a bad conversation, and improving the agent. What each one needs recorded.

How an AI Support Agent Stays Grounded in Your Help Center and Policies

Grounding an AI support agent: a relevance floor on retrieval, policy kept consistent with live order data, and why content quality decides the answer.

Multilingual AI Support: One Agent Across Languages vs One Per Market

A multilingual AI support agent replies in many languages from one model - but the knowledge base, per-language quality, and market rules are what fork.