Arun ArumugamAgent Engineering

The Feed

Blogs from the Feed, newest first.

RSSJSON

  1. Anthropic EngineeringBlogInfraRedesigning source control at AnthropicMerges grew more than tenfold as agents took on more work, so a seven-person team replaced Git with an in-house version control system in five months.
  2. EarendilBlogAgentsPi Durable on CloudflareA worked build of long-running pi agents: one session per Durable Object with SQLite storage, hibernation when idle, alarm wakeups and a search subagent.
  3. VercelBlogSecurityFindings from the $1M Vercel Sandbox challengeAn open challenge to break out of Vercel's agent sandbox drew 1,285 reports, found a Linux kernel zero-day and paid $242K. Real data on how well agent isolation holds up.

  1. Spotify EngineeringBlogInfraAdvancing generative recommendation inference with NVIDIA TensorRT-LLMHow Spotify serves LLM-style generative recommenders in production with TensorRT-LLM. Inference optimization beyond chatbots.
  2. Netflix Tech BlogBlogOntology-driven observability: an end-to-end knowledge graph at Netflix scaleNetflix describes how it modelled its infrastructure as an ontology-backed knowledge graph so engineers and GenAI tools can trace incidents across services handling billions of requests a day.
  3. Google ResearchBlogGrounding AI agents in biological reality: building blocks for autonomous wet labsGoogle Research lays out the pieces needed for AI agents to run real lab experiments with robotics, and how it keeps agent plans tied to what actually happens at the bench.
  4. Hugging FaceBlogScaling Indic pretraining: 2.8 trillion tokens and an open data generation pipelineNVIDIA explains how it built a 2.8T-token pretraining corpus for Indian languages and open-sourced the synthetic data generation pipeline behind it.

  1. Meta EngineeringBlogAgentsHow Meta built Forum with AI-native engineeringHow a Meta team built a new app with AI agents in its engineering workflow from day one. Practice at real scale, not a demo.
  2. VercelBlogModelsOpen weight takes 71.5% of tokens and 23% of spendProduction data from Vercel's AI Gateway for September: open-weight models handled over 70% of tokens, AI spend grew 13x and cost per token halved.

  1. Netflix Tech BlogBlogResearchThe lifecycle of LLM-as-a-judge: building, aligning and monitoring at scaleNetflix covers running LLM judges in production, from building them to aligning them with human judgment and monitoring them over time. Useful for anyone evaluating agents.

  1. CloudflareBlogInfraHow D1 scales to millions of databasesHow Cloudflare's D1 control plane manages millions of databases while queries stay at the edge, with design principles that carry over to other systems.
  2. LangChainBlogAgentsImproving Deep Agents with harness engineeringStarting from 52.8% with GPT-5.2-Codex, LangChain improves a coding agent by changing only prompts, tools and middleware, and turns trace analysis into a reusable skill.

  1. Meta EngineeringBlogProtocolsShipping Media over QUIC at Meta scale for Instagram Live ingestMeta is moving Instagram Live ingest to Media over QUIC, an emerging open protocol, to simplify its live-streaming stack. A rare production rollout of the protocol.

  1. Ramp EngineeringBlogInfraPutting a transformer inside every card authorizationRamp runs a transformer over each cardholder's transaction history on every authorization in real time, ranking fraud 14% better than its old model.