What is mattwood.fyi?

A live, structured feed of riffs and links from Matt Wood, Chief AI & Technology Officer at AWS — readable by people and queryable by agents as a typed knowledge graph.

Is there an API and is it free?

Yes. A public, read-only query API is available with no authentication and no API keys. See https://mattwood.fyi/openapi.json and https://mattwood.fyi/docs.

How should an AI agent use this site?

Fetch https://mattwood.fyi/feed.json for the full structured feed, or query the API for keyword/semantic search and the connection graph. See https://mattwood.fyi/agents/ for when-to-use guidance.

Training a 4B Model for Faster SQL Query Plans

Connections

Contrastive Language Model (CLM) Both involve specialized training approaches for language models — contrastive learning for representation quality and fine-tuning for SQL query plan optimization
Parseable: Observability Infrastructure Platform Both involve high-performance SQL query processing on columnar data at scale — Parseable uses columnar storage with SQL/PromQL for telemetry, while the 4B model targets faster SQL query plans
GenRec: LLM-Native Recommendation System at Netflix Both represent industry cases of training specialized models (4B for SQL, LLM for recommendations) for production systems with performance constraints, reflecting the trend of task-specific LLM deployment
Real-Time Video Generation on Trainium Both focus on kernel-level and hardware-specific optimizations to accelerate AI model execution — one for video generation on Trainium, the other for SQL query planning on a 4B model
Dust: Pretraining Transformers Without Backpropagation Both explore non-standard optimization strategies to improve training efficiency for specific model sizes and tasks
Engineering a Calibrated Classifier Model Both describe fine-tuning smaller/specialized models (4B SQL model vs 2B Hobson classifier) optimized for specific tasks with latency/performance tradeoffs rather than using frontier models
PLUM: Adapting Pre-trained Language Models for Industrial-scale Generative Recommendations Both involve adapting/fine-tuning models on domain-specific data for specialized industrial tasks (SQL query plans vs. recommendations), sharing the continued pre-training and domain adaptation methodology

Type:	link
First published:	Wednesday, 16 September 2026 at 23:11 UTC
Location:	https://rohanbansal.com/qorl
URL:	https://mattwood.fyi/i/5a8c8bb3

I'm Matt Wood, and this is For Your Information. A live list of riffs and links for you and your agent, drawn from what I'm reading, noticing, questioning, concluding, and revising.

Connections