Research is a pipeline: search discovers, fetch reads
A server-side research workflow should separate source discovery from source reading, preserve the URLs and extracted content as evidence, and evaluate each stage independently.
Read postAIVAX / Blog
Notes from the operating layer.
Field notes on AI infrastructure, retrieval, model operations and the systems that keep every request visible.
14 published field notesLatest blog posts
-
-
Give your coding agent a memory it can read and a world it can fetch
Coding agents lose context between sessions and guess at the world outside the repo. AIVAX's Collections and Web Utilities MCP servers give them writable semantic memory plus fetch and search — with scoped credentials, bounded retrieval, and explicit write control.
Read post -
Persistent memory is a write path, not a notebook
An agent that writes its own memory gains a second input channel that outlives the session. Recent poisoning research shows why that channel needs ownership, expiry, and review — and how AIVAX scopes it.
Read post -
Batch is an admission-control problem, not a queue
Bulk AI work fails at the boundary between the queue and the provider: rate limits, balance, validation, and overload. AIVAX Batch answers with bounded admission, per-item validation, and failure-shaped retries.
Read post -
From test to monitor: the judge trajectory as a production signal
Agentic Tests already scores every turn of a simulated conversation. Applied to real production traffic, the same trajectory signal — score, at-risk state, persistent loss — tells you when a live conversation is drifting before the user gives up.
Read post -
MCP went stateless. Your application did not
MCP 2026-07-28 removes protocol sessions and transport replay, moving durable state, retries, long-running work, and compatibility into explicit application contracts.
Read post -
A vector database is not a RAG system
A vector database stores and searches vectorized records; a managed RAG system coordinates the wider path from source content to maintained, retrievable context.
Read post -
An MCP server is a trust boundary, not just a tool catalog
Connecting an MCP server exposes executable capabilities to a model. Discovery policy, call-time authorization, result handling, and audit decide whether the connection is safe — not the tool list.
Read post -
Migrating from OpenAI Assistants Without Rebuilding the Same Coupling
The Assistants API shutdown is an architecture migration, not an endpoint rename. Map configuration, state, execution, tools and observability before choosing Responses or an application-owned gateway design.
Read post -
Reasoning Is a Protocol, Not Just a Model Setting
Reasoning APIs now carry summaries, signatures, encrypted state and provider-specific stream events. A production gateway must preserve that protocol, not flatten it into text.
Read post -
Gateway Moderation Is a Product Control, Not Just a Prompt
AIVAX Gateway Moderation centralizes safety policy, normally evaluating incoming content before model generation while preserving a configured fallback when that separate step is unavailable.
Read post -
Introducing Agentic Tests
Agentic Tests evaluates complete, goal-oriented conversations through your AI Gateway, with a simulated user and an independent judge tracking the outcome across turns.
Read post -
Semantic Search vs. Reranking: They Solve Different Problems
A practical guide to diagnosing retrieval, rank fusion, and reranking—so you improve the stage that is actually failing.
Read post -
Reflex: retrieval built for recurring documents
Reflex combines semantic relevance, lexical evidence and account-scoped cache reuse for recurring-document retrieval.
Read post