DeepSeek open-sourced an agent harness where the model, tools, memory, and even the loop itself are swappable plugins. A hands-on look at the pattern and how to apply it to your own agent stack.
Daily English practice for tech professionals. Morning session covering backpressure, dead letter queue, control plane, canary deployment, and tail latency with pronunciation guide, exercises, and real-world examples.
OpenAI just cut GPT-5.6 Luna pricing by 80% and Terra by 20%. Combined with 15% better token efficiency from speculative decoding, the cost math for AI products has fundamentally shifted. Here's how to rethink your model routing strategy and capture the savings without sacrificing quality where it matters.
Daily English practice for tech professionals. Morning session covering bulkhead, blast radius, circuit breaker, idempotent, and throughput — the reliability pattern vocabulary every backend engineer needs to discuss system design in English with confidence.
APIs designed for human developers fail in predictable ways when AI agents consume them. After shipping several agent-integrated systems in production, here are the concrete patterns — idempotency, rich errors, context anchors, and MCP alignment — that determine whether your API is agent-ready or a liability.
A Stanford, SambaNova, and UC Berkeley paper introduces Agentic Context Engineering (ACE) — a Generator-Reflector-Curator split that delivers +10.6% improvement on agent tasks and +8.6% on finance tasks, without any gradient updates. Here's what the architecture means for teams building production AI agents.
After PocketOS lost their production database in 9 seconds to a Cursor agent, and Replit destroyed 1,200+ executive records, the industry has enough post-mortems to see patterns. Here are the 7 failure classes, what each looks like in the logs, and concrete guardrail architecture for each.
Claude Opus 5 isn't just a better model — it changes the economics and architecture of AI agent systems. What the benchmarks actually mean for teams building production pipelines: configurable effort, self-verification, and the 'chief-of-staff' pattern.
The July 28 MCP specification eliminates sessions, removes the initialize handshake, and makes every request self-contained. A technical breakdown of what changed, why it matters for load balancers and multi-tenant deployments, and how to migrate existing servers.
Isolation is the easy part. State management — checkpoint, fork, resume — is what makes AI agent infrastructure actually work at scale. How Perplexity built SPACE in 10 weeks, hit 1.25M sandbox creations at launch, and cut P90 latency from 447ms to 89ms.
Daily English practice for tech professionals. Morning session covering architecture vocabulary — throughput, bottleneck, horizontal scaling, fault tolerance, load balancing — with pronunciation guide, exercises, and real-world tech examples.
Your 100-step AI agent crashes at step 87. What happens next? Durable execution solves the fundamental fault-tolerance problem in production agentic workflows. Here's the technical breakdown and decision framework.
Daily English practice for tech professionals. Morning session covering architecture vocabulary — idempotent APIs, fault tolerance, service mesh — with pronunciation guide and exercises.
Stateless agents are demos. Stateful agents are products. Here's the memory architecture stack that separates a useful AI agent from one that forgets everything between calls — and how to build it without rebuilding your whole infrastructure.
GPT-5.6, Claude Sonnet 5, and Grok 4.5 all launched within 10 days of each other in July 2026. No single model wins every workload. Here's how to build a routing layer that cuts your AI bill by 40-80% without sacrificing quality.
Most AI cost overruns aren't about model pricing — they're about bloated context windows. Here's the architecture that helped teams cut AI costs by 80% without sacrificing output quality.
AWS, Microsoft, and Google are all rebuilding agent runtimes around the session — not the request. Here's what this architectural shift means for teams building production AI systems.
Shadowing and pronunciation drills for explaining architectural decisions to your team. Includes real decision scripts, hedging language, and a 5-minute confidence drill for Vietnamese developers.
Prompt engineering gets you to demo. Context engineering gets you to production. Learn how controlling what information each agent receives determines whether your agentic system is reliable or brittle.
Daily English practice for tech professionals. Evening session — speaking practice, review, and real-world scenarios. Wednesday review: all architecture vocabulary from today + communication techniques for explaining complex systems to anyone.
Daily English practice for tech professionals. Morning session covering architecture vocabulary — idempotent, scalability, fault tolerance, eventual consistency — with pronunciation guide and real-world examples.
Daily English practice for tech professionals. Noon session architecture vocabulary deep dive: circuit breaker, sharding, CQRS, read replica, service mesh. Pronunciation guide and exercises.
A complete shadowing script for explaining how your system works to a new team member. Read it aloud three times, then say it in your own words. Includes pronunciation drills for th, v/f, and consonant cluster sounds.
How to lead architecture discussions in English — phrases for proposing designs, handling pushback, building consensus, and making decisions with international teams.
Traditional OAuth and service accounts don't work for multi-agent AI systems. Here's what Uber and Auth0 discovered — and how to design agent identity the right way.
A deep technical dive into Atlassian's cloud migration journey — stateless compute, Tenant Context Service, CQRS, and why they never rewrote Jira.
A practical guide to building an AI-powered monitoring dashboard that shows EC2 health, web platform status, CI/CD quality checks, and uses Claude to interpret anomalies in plain English.
After deploying multi-agent AI systems at scale, here are the failure patterns we hit and the resilience patterns that saved us. Practical guide for engineering teams.
Daily English practice for tech professionals. Evening session — architecture vocab review, speaking practice, and how to explain complex systems to non-technical stakeholders.
Daily English practice for tech professionals. Morning session covering system design and cloud architecture vocabulary with pronunciation guide, exercises, and real-world examples.
Daily English practice for tech professionals. Noon session — architecture vocabulary deep dive with pronunciation, exercises, and real-world examples.
I spent 3 hours reading Hermes Agent source code — and discovered why most AI agents never improve. A deep dive into its 8-loop temporal architecture that makes AI compound over time.
A technical deep dive into Hermes Agent's 8-loop temporal architecture — how each loop works, what triggers it, and how they compound to create an AI system that improves with use.
How to design a Hermes Agent coding workflow that learns your codebase, accumulates reusable patterns, and accelerates feature development over time.
Step-by-step installation guide for Hermes Agent — from prerequisites to first skill creation, with troubleshooting tips and production-ready configuration examples.
How to build an automated quality control agent using Hermes — from code review automation to regression detection, test coverage analysis, and continuous quality improvement.
How to design an effective skill tagging and retrieval architecture in Hermes Agent — taxonomy design, tag hierarchies, retrieval optimization, and real-world tagging examples.
A Tech Lead's guide to evaluating, piloting, and scaling Hermes Agent adoption across an engineering team — with ROI calculations, change management tips, and common failure modes.
Daily English practice for tech professionals. Evening session — review architecture vocabulary, learn to explain complex systems in plain English, and practice speaking with technical stakeholders.
Daily English practice for tech professionals. Morning session covering architecture vocabulary — microservices, distributed systems, cloud-native design with pronunciation and real-world examples.
Daily English practice for tech professionals. Noon session — architecture vocabulary deep dive: scalability, latency, microservices, and how to explain system design to non-tech stakeholders.
Is your Claude Code token usage insanely high? The culprit might be your architecture. Learn how DRY principles and clear system design slash AI token costs.
Microsoft cancelled thousands of Claude Code licenses. Uber burned its entire 2026 AI budget in four months. The technical reason why - and what every engineering team should do about it.
A Tech Lead's field notes from shipping production AI agents — why custom agents still matter, how to avoid the infra money pit, the architecture pattern that actually works, and why understanding your client matters more than understanding transformers.
How to architect a multi-agent system using Domain-Driven Design principles. Define bounded contexts, domain events, state machines, and communication patterns for your AI software team.
A complete technical guide to building a profitable agentic AI system using only open-source tools — with retrieval, orchestration, tool use, and observability. Includes architecture diagrams and real cost analysis.
Anthropic's Model Context Protocol crossed 97M installs in March 2026. As a Technical Lead who's been building with MCP since early days, here's what that milestone actually means — and why every developer building AI systems needs to understand it now.
Meta dropped Llama 4 Scout, Maverick, and Behemoth. Google fired back with Gemma 4. As a Technical Lead, here's what these releases actually mean for your teams and projects.
Alibaba just released its third proprietary model in days. Google's Gemini Flash-Lite costs $0.25 per million tokens. NVIDIA's Nemotron runs 2.2x faster than GPT-OSS-120B. The LLM cost war has arrived — here's what it means for architects choosing AI infrastructure in 2026.
The A2A protocol under the Linux Foundation is quietly becoming the HTTP of the agentic era. Here's what it means for enterprise architects, why it matters more than another model release, and how to think about it from a systems design perspective.
Daily English practice for tech professionals. Morning session covering system architecture vocabulary with pronunciation guide, exercises, and real-world examples.
Practical English phrases for Vietnamese tech leads navigating technical disagreements, design discussions, and architecture decision meetings with international teams.
The Model Context Protocol crossed 10,000 published servers under the Linux Foundation's Agentic AI Foundation. As someone who's integrated dozens of AI systems, here's why this number matters more than any benchmark.
GPT-4 level AI cost $30/M tokens in 2023. Today it's under $1. Here's the technical architecture that lets you capture 90%+ of that savings without sacrificing quality.
A comprehensive guide to the modern free-tier tech stack that lets you build, deploy, and scale a startup for roughly $20/month. No servers. No DevOps team. No funding required. Just an idea and WiFi.
90% of developers now use AI at work. But the real shift in March 2026 is agents moving from suggestion-mode to autonomous execution. Here's what that actually looks like in production systems and what breaks when you go too far too fast.
Daily English practice for tech professionals. Morning session covering System Architecture vocabulary — microservices, scalability, fault tolerance, load balancing, and cloud design patterns — with pronunciation guide, exercises, and real-world examples.
Daily English practice for tech professionals. Noon session — vocabulary deep dive with pronunciation, exercises, and real-world examples.
Gartner says 40% of enterprise apps will embed AI agents this year. But 40% of agentic projects will be scrapped by 2027. Here's what separates the teams that ship production agents from those that get stuck in pilots forever.
Practical English phrases and dialogue templates for Vietnamese Tech Leads running architecture reviews, proposing solutions, and handling technical disagreements in international teams.
The exact language patterns senior engineers use to propose technical solutions, run architecture review meetings, push back on bad ideas without damaging relationships, and write ADRs that actually get read and followed.
Anthropic's Agent Teams feature in Claude Opus 4.6 lets multiple Claude Code instances work in parallel on the same codebase. Here's the architectural model, real-world performance data, and what actually changes for teams building production software.
How AI is reshaping software architecture from the inside — ML-powered pattern selection, LLM orchestration, AI gateways, event-driven agents, and a full team implementation roadmap with diagrams and production-ready code.
GPT-5.4, Gemini 3.1 Pro, and Claude 4.6 are now neck-and-neck on benchmarks. When the models are equal, everything else becomes the differentiator. Here's how to choose.
When you're managing 10, 20, or 50 Umbraco sites, individual project economics don't work. The Marketing OS framework: shared NuGet packages, shared document type libraries, AI-accelerated delivery, and how to reduce per-site migration cost by 50–70%.
AI can generate impressive code that completely ignores your architecture. How to provide architectural context, enforce patterns, and know when to override AI suggestions.
Architecture buzzwords demolished and rebuilt. Two tech leads explain when event-driven architecture, CQRS, and the Saga pattern are genuinely useful — and when they're just resume padding. Pizza delivery analogies included.
Most developers skip straight to 'generate code.' The teams that get real value from AI spend 40% of their time planning before writing a single prompt.
Not every team wants React. Here are five production-ready alternatives for building marketing websites with headless Umbraco 17 — from Astro and Nuxt to SvelteKit, .NET Razor, and even plain HTML.
Turning MarketingOS into a reusable template: new client onboarding in under an hour, multi-tenant content management, cost analysis showing 70% reduction per site, lessons learned, and what I'd do differently.
PostgreSQL, MongoDB, Redis, DynamoDB — when do you use what? Two tech leads break it down with filing cabinet analogies, real use cases, and zero religious wars.
Putting AI into production is nothing like building a demo. Two tech leads discuss costs, hallucinations, latency, guard rails, and what actually breaks when real users hit your AI features.
The real cost of AI voice interviews, broken down per minute. Managed vs self-hosted economics, the three tipping points, and how to get from $3.45 per interview to under $1.00.
From 10 concurrent interviews to 10,000. LiveKit SFU mesh, stateless agent workers, Kubernetes auto-scaling, regional deployment, and the infrastructure patterns that handle hiring season surges.
Two tech leads break down API design styles using restaurant analogies. Learn when to use REST, GraphQL, or gRPC — explained in casual English perfect for listening practice.
The framework debate almost split the team. Flutter, React Native, or going fully native? Here's our decision matrix and what we learned comparing all three.
A casual conversation about microservices vs monoliths. Two tech leads explain when to split, when to stay, and why most teams get it wrong — in plain English you can listen to and learn from.
Why headless Umbraco 17 with Next.js is the sweet spot for reusable marketing websites, the architecture decisions behind MarketingOS, and setting up both projects with Clean Architecture on .NET 10.
Why research interviews need server-side voice agents, the three-tier architecture, room metadata as configuration transport, and the 100-500ms propagation latency nobody tells you about.
Setting up an Angular 21 ecommerce project from scratch. Nx monorepo, feature-based folder structure, ESLint flat config, Vitest, and auto-generated Kiota TypeScript API client from .NET 10.
A Technical Lead's honest assessment of Angular 21 (released Nov 2025), .NET 10 LTS, and GitHub Copilot for an Ecommerce project. Architecture decisions, ADR templates, and what nobody tells you on day one.
The Tech Lead's honest retrospective on Angular 21 + .NET 10 for an ecommerce project. What this stack does well, where it struggles, the risk register, and practical advice for teams starting this journey.
Post A — How to enforce Nx module boundary rules in Angular 21 to prevent spaghetti imports, protect domain separation, and keep teams independently productive in a monorepo.
Taking Kids Learn to production with Docker, CI/CD, OpenTelemetry observability, performance optimization, Native AOT considerations, and an honest retrospective on what worked and what was over-engineered.
How AI transforms the Solution Architect role: AI-assisted architecture diagramming, ADR generation, trade-off analysis, technology selection rationale, and the architectural taste that only experience provides.
Testing strategy per layer for Kids Learn. Domain unit tests without mocks, Application tests with NSubstitute, integration tests with Testcontainers, and NetArchTest for enforcing the Dependency Rule.
Why pure Clean Architecture scatters features across projects, how Vertical Slice Architecture solves this, and the hybrid approach we use in Kids Learn.
Implementing Infrastructure with EF Core 10 (pgvector, JSON columns), AI service integrations, Minimal APIs with Route Groups, authentication, and wiring it all together in Program.cs.
The cascaded STT→LLM→TTS pipeline gives you control. Speech-to-speech models give you speed. Here's how to choose — and why the best systems use both.
Command/Query separation for Kids Learn, pipeline behaviors, the MediatR licensing debate, Wolverine as the modern alternative, and the Repository Pattern in 2026.
The landscape of real-time voice AI has shifted. Gemini Live, OpenAI Realtime, Bedrock Nova Sonic, and Grok make sub-500ms AI conversations possible. Here's the reference architecture for building a production voice interview platform.
Rich domain models for Kids Learn with C# 14. Entities with invariants, value objects, domain events, aggregate roots, and architecture tests to enforce the rules.
What Clean Architecture actually means in .NET 10, the Dependency Rule explained with real code, C# 14 features that matter, and setting up the Kids Learn solution structure.
How I designed, analyzed, implemented, and tested Kids Learn — an AI-powered educational SaaS platform — using Claude as my development partner, Gemini for AI features, Next.js, PostgreSQL, and pgvector. A complete walkthrough from napkin sketch to production.
After trying Next.js, Gatsby, Hugo, and plain HTML for personal sites, I finally found the framework that fits. Here's why Astro won.
The story of migrating a 2M+ line retail platform to microservices. What the strangler pattern actually looks like when you're in the middle of it.
Get notified when I publish new posts on AI, .NET, cloud architecture and more.
