<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>Saram Learn | Engineering &amp; Curiosity Knowledge Base</title><description>Technical notes, architectural blueprints, and curiosity deep-dives on AI Agents, Edge Computing, and Regulated Life Sciences.</description><link>https://learn.saram.io/</link><language>en-us</language><item><title>The State of Lab Automation in 2026: An Exhaustive Technical Deep Dive</title><link>https://learn.saram.io/notes/state-of-lab-automation-deep-dive-2026/</link><guid isPermaLink="true">https://learn.saram.io/notes/state-of-lab-automation-deep-dive-2026/</guid><description>A comprehensive analysis of lab automation in life sciences, biotech, and biopharma: liquid handling physics, robotic workcells, schedulers, SiLA 2 vs OPC UA LADS, and closed-loop self-driving labs.</description><pubDate>Thu, 08 Oct 2026 00:00:00 GMT</pubDate><category>lab-informatics</category><category>Lab Automation</category><category>Biotech</category><category>Liquid Handling</category><category>SiLA 2</category><category>OPC UA</category><category>Robotics</category><category>LIMS</category><category>Hardware</category></item><item><title>Why Cloudflare Workers (Not Pages) is the True Modern Edge Architecture</title><link>https://learn.saram.io/notes/cloudflare-workers-vs-pages-architecture/</link><guid isPermaLink="true">https://learn.saram.io/notes/cloudflare-workers-vs-pages-architecture/</guid><description>A deep dive into why Cloudflare Workers with Static Assets has superseded Cloudflare Pages, how V8 isolates operate, and why modern Astro projects build on Workers.</description><pubDate>Thu, 01 Oct 2026 00:00:00 GMT</pubDate><category>edge-systems</category><category>Cloudflare Workers</category><category>Edge Computing</category><category>Astro</category><category>Architecture</category><category>Infrastructure</category></item><item><title>Deconstructing Multi-Agent Consensus: Evaluator-Optimizer Loops vs Linear Chains</title><link>https://learn.saram.io/notes/multi-agent-consensus-loops-2026/</link><guid isPermaLink="true">https://learn.saram.io/notes/multi-agent-consensus-loops-2026/</guid><description>Why monolithic prompt chains degrade on complex tasks, how iterative evaluator-optimizer topologies converge, and practical patterns for production multi-agent systems.</description><pubDate>Mon, 28 Sep 2026 00:00:00 GMT</pubDate><category>ai-agents</category><category>AI Agents</category><category>Orchestration</category><category>System Design</category><category>Evaluation</category><category>DSPy</category></item><item><title>The Three-Database Architecture for Hybrid GraphRAG: Relational, Vector &amp; Graph</title><link>https://learn.saram.io/notes/hybrid-graphrag-three-database-architecture/</link><guid isPermaLink="true">https://learn.saram.io/notes/hybrid-graphrag-three-database-architecture/</guid><description>Why single-vector similarity search fails on enterprise knowledge graphs, and how unifying Postgres/SQLite, Qdrant, and Neo4j eliminates hallucinated relationships.</description><pubDate>Sun, 20 Sep 2026 00:00:00 GMT</pubDate><category>graphrag</category><category>GraphRAG</category><category>Knowledge Graphs</category><category>Vector Search</category><category>Qdrant</category><category>Databases</category></item><item><title>Validating the Unpredictable: Computer Software Assurance (CSA) for AI in GxP</title><link>https://learn.saram.io/notes/validating-probabilistic-ai-in-gxp/</link><guid isPermaLink="true">https://learn.saram.io/notes/validating-probabilistic-ai-in-gxp/</guid><description>How life sciences organizations reconcile FDA 21 CFR Part 11, GAMP 5, and QMSR with non-deterministic LLMs through statistical validation harnesses.</description><pubDate>Sat, 12 Sep 2026 00:00:00 GMT</pubDate><category>gxp-ai</category><category>GxP Compliance</category><category>Life Sciences</category><category>FDA</category><category>CSA</category><category>ISO 42001</category><category>Validation</category></item><item><title>Multi-Token Prediction (MTP) &amp; Speculative Decoding in Local LLM Inference</title><link>https://learn.saram.io/notes/multi-token-prediction-and-local-inference/</link><guid isPermaLink="true">https://learn.saram.io/notes/multi-token-prediction-and-local-inference/</guid><description>How breaking the single-token autoregressive bottleneck unlocks 2.5x speedups on edge hardware and local workstations.</description><pubDate>Sat, 29 Aug 2026 00:00:00 GMT</pubDate><category>local-llm</category><category>Local LLM</category><category>Inference</category><category>MTP</category><category>Speculative Decoding</category><category>Hardware</category></item><item><title>Hardening Model Context Protocol (MCP) in Production &amp; Regulated Environments</title><link>https://learn.saram.io/notes/model-context-protocol-in-production/</link><guid isPermaLink="true">https://learn.saram.io/notes/model-context-protocol-in-production/</guid><description>How to architect secure, audited MCP gateways for enterprise agent swarms: authorization scopes, rate limiting, and 21 CFR Part 11 compliance.</description><pubDate>Fri, 14 Aug 2026 00:00:00 GMT</pubDate><category>ai-agents</category><category>MCP</category><category>Model Context Protocol</category><category>Security</category><category>AI Agents</category><category>Architecture</category></item><item><title>Building Low-Latency Scientific Instrument Ingestion Pipelines for LIMS</title><link>https://learn.saram.io/notes/modern-lab-instrument-iot-pipelines/</link><guid isPermaLink="true">https://learn.saram.io/notes/modern-lab-instrument-iot-pipelines/</guid><description>How to connect physical chromatography and mass spectrometry systems to modern cloud/edge LIMS using OPC UA, edge gateways, and immutable audit trails.</description><pubDate>Wed, 22 Jul 2026 00:00:00 GMT</pubDate><category>lab-informatics</category><category>LIMS</category><category>Laboratory Automation</category><category>IoT</category><category>OPC UA</category><category>Life Sciences</category><category>Edge</category></item></channel></rss>