Writing & Engineering Notes
AI systems, shipped clearly.
Notes on agentic workflows, RAG systems, ML engineering, and building useful AI products. Deep-dives into agentic workflows, RAG, and applied ML.
12 entries found
Updated guide to parameter-efficient fine-tuning, covering recent advances in low-rank adaptation, quantization, multi-task adaptation, and hardware-aware optimizations that make customizing large models accessible on consumer hardware.
Analysis of MetaComp's StableX Know Your Agent (KYA) Framework—a governance standard for AI agents operating in payments, compliance, and wealth management, launched at Money20/20 Asia.
Deep dive into Microsoft's open‑source Agent Governance Toolkit—a hypervisor‑based framework that brings deterministic policy enforcement, zero‑trust identity, and execution sandboxing to autonomous AI agents.
Deep dive into Hugging Face's ml‑intern—an open‑source AI agent that automates end‑to‑end LLM post‑training workflows, from literature review and data validation to fine‑tuning and deployment.
Deep dive into OpenAI's newly released Agents SDK—a lightweight, production‑ready Python framework for orchestrating multi‑agent workflows with built‑in tool‑calling, memory management, and real‑time streaming.
A technical deep dive into Google's open‑source library for building composable, asynchronous AI pipelines — with 2,108 stars, 212 forks, and a dual‑interface pattern that abstracts away streaming complexity.